id sid tid token lemma pos 12483 1 1 Automated Automated NNP 12483 1 2 Fake Fake NNP 12483 1 3 News News NNP 12483 1 4 Detection Detection NNP 12483 1 5 in in IN 12483 1 6 the the DT 12483 1 7 Age Age NNP 12483 1 8 of of IN 12483 1 9 Digital Digital NNP 12483 1 10 Libraries Libraries NNPS 12483 1 11 ARTICLE ARTICLE NNP 12483 1 12 Automated Automated NNP 12483 1 13 Fake Fake NNP 12483 1 14 News News NNP 12483 1 15 Detection Detection NNP 12483 1 16 in in IN 12483 1 17 the the DT 12483 1 18 Age Age NNP 12483 1 19 of of IN 12483 1 20 Digital Digital NNP 12483 1 21 Libraries Libraries NNP 12483 1 22 Uğur Uğur NNP 12483 1 23 Mertoğlu Mertoğlu NNP 12483 1 24 and and CC 12483 1 25 Burkay Burkay NNP 12483 1 26 Genç Genç NNP 12483 1 27 INFORMATION INFORMATION VBD 12483 1 28 TECHNOLOGY technology NN 12483 1 29 AND and CC 12483 1 30 LIBRARIES library NNS 12483 1 31 | | NNS 12483 1 32 DECEMBER DECEMBER NNS 12483 1 33 2020 2020 CD 12483 1 34 https://doi.org/10.6017/ital.v39i4.12483 https://doi.org/10.6017/ital.v39i4.12483 NNP 12483 1 35 Uğur Uğur NNP 12483 1 36 Mertoğlu Mertoğlu NNP 12483 1 37 ( ( -LRB- 12483 1 38 umertoglu@hacettepe.edu.tr umertoglu@hacettepe.edu.tr NNP 12483 1 39 ) ) -RRB- 12483 1 40 is be VBZ 12483 1 41 a a DT 12483 1 42 PhD phd NN 12483 1 43 Candidate Candidate NNP 12483 1 44 , , , 12483 1 45 Hacettepe Hacettepe NNP 12483 1 46 University University NNP 12483 1 47 . . . 12483 2 1 Burkay Burkay NNP 12483 2 2 Genç Genç NNP 12483 2 3 ( ( -LRB- 12483 2 4 bgenc@cs.hacettepe.edu.tr bgenc@cs.hacettepe.edu.tr NNP 12483 2 5 ) ) -RRB- 12483 2 6 is be VBZ 12483 2 7 Assistant Assistant NNP 12483 2 8 Professor Professor NNP 12483 2 9 , , , 12483 2 10 Hacettepe Hacettepe NNP 12483 2 11 University University NNP 12483 2 12 . . . 12483 3 1 © © NNP 12483 3 2 2020 2020 CD 12483 3 3 . . . 12483 4 1 ABSTRACT ABSTRACT NNP 12483 4 2 The the DT 12483 4 3 transformation transformation NN 12483 4 4 of of IN 12483 4 5 printed print VBN 12483 4 6 media medium NNS 12483 4 7 into into IN 12483 4 8 the the DT 12483 4 9 digital digital JJ 12483 4 10 environment environment NN 12483 4 11 and and CC 12483 4 12 the the DT 12483 4 13 extensive extensive JJ 12483 4 14 use use NN 12483 4 15 of of IN 12483 4 16 social social JJ 12483 4 17 media medium NNS 12483 4 18 have have VBP 12483 4 19 changed change VBN 12483 4 20 the the DT 12483 4 21 concept concept NN 12483 4 22 of of IN 12483 4 23 media medium NNS 12483 4 24 literacy literacy NN 12483 4 25 and and CC 12483 4 26 people people NNS 12483 4 27 ’s ’s POS 12483 4 28 habits habit NNS 12483 4 29 of of IN 12483 4 30 news news NN 12483 4 31 consumption consumption NN 12483 4 32 . . . 12483 5 1 While while IN 12483 5 2 online online JJ 12483 5 3 news news NN 12483 5 4 is be VBZ 12483 5 5 faster fast JJR 12483 5 6 , , , 12483 5 7 easier easy JJR 12483 5 8 , , , 12483 5 9 comparatively comparatively RB 12483 5 10 cheaper cheap JJR 12483 5 11 , , , 12483 5 12 and and CC 12483 5 13 offers offer VBZ 12483 5 14 convenience convenience NN 12483 5 15 in in IN 12483 5 16 terms term NNS 12483 5 17 of of IN 12483 5 18 people people NNS 12483 5 19 's 's POS 12483 5 20 access access NN 12483 5 21 to to IN 12483 5 22 information information NN 12483 5 23 , , , 12483 5 24 it -PRON- PRP 12483 5 25 speeds speed VBZ 12483 5 26 up up RP 12483 5 27 the the DT 12483 5 28 dissemination dissemination NN 12483 5 29 of of IN 12483 5 30 fake fake JJ 12483 5 31 news news NN 12483 5 32 . . . 12483 6 1 Due due IN 12483 6 2 to to IN 12483 6 3 the the DT 12483 6 4 free free JJ 12483 6 5 production production NN 12483 6 6 and and CC 12483 6 7 consumption consumption NN 12483 6 8 of of IN 12483 6 9 large large JJ 12483 6 10 amounts amount NNS 12483 6 11 of of IN 12483 6 12 data datum NNS 12483 6 13 , , , 12483 6 14 fact fact NN 12483 6 15 - - HYPH 12483 6 16 checking check VBG 12483 6 17 systems system NNS 12483 6 18 powered power VBN 12483 6 19 by by IN 12483 6 20 human human JJ 12483 6 21 efforts effort NNS 12483 6 22 are be VBP 12483 6 23 not not RB 12483 6 24 enough enough JJ 12483 6 25 to to TO 12483 6 26 question question VB 12483 6 27 the the DT 12483 6 28 credibility credibility NN 12483 6 29 of of IN 12483 6 30 the the DT 12483 6 31 information information NN 12483 6 32 provided provide VBN 12483 6 33 , , , 12483 6 34 or or CC 12483 6 35 to to TO 12483 6 36 prevent prevent VB 12483 6 37 its -PRON- PRP$ 12483 6 38 rapid rapid JJ 12483 6 39 dissemination dissemination NN 12483 6 40 like like IN 12483 6 41 a a DT 12483 6 42 virus virus NN 12483 6 43 . . . 12483 7 1 Libraries library NNS 12483 7 2 , , , 12483 7 3 long long RB 12483 7 4 known know VBN 12483 7 5 as as IN 12483 7 6 sources source NNS 12483 7 7 of of IN 12483 7 8 trusted trusted JJ 12483 7 9 information information NN 12483 7 10 , , , 12483 7 11 are be VBP 12483 7 12 facing face VBG 12483 7 13 challenges challenge NNS 12483 7 14 caused cause VBN 12483 7 15 by by IN 12483 7 16 misinformation misinformation NN 12483 7 17 as as IN 12483 7 18 mentioned mention VBN 12483 7 19 in in IN 12483 7 20 studies study NNS 12483 7 21 about about IN 12483 7 22 fake fake JJ 12483 7 23 news news NN 12483 7 24 and and CC 12483 7 25 libraries.1 libraries.1 CD 12483 7 26 Considering consider VBG 12483 7 27 that that IN 12483 7 28 libraries library NNS 12483 7 29 are be VBP 12483 7 30 undergoing undergo VBG 12483 7 31 digitization digitization NN 12483 7 32 processes process NNS 12483 7 33 all all RB 12483 7 34 over over IN 12483 7 35 the the DT 12483 7 36 world world NN 12483 7 37 and and CC 12483 7 38 are be VBP 12483 7 39 providing provide VBG 12483 7 40 digital digital JJ 12483 7 41 media medium NNS 12483 7 42 to to IN 12483 7 43 their -PRON- PRP$ 12483 7 44 users user NNS 12483 7 45 , , , 12483 7 46 it -PRON- PRP 12483 7 47 is be VBZ 12483 7 48 very very RB 12483 7 49 likely likely JJ 12483 7 50 that that IN 12483 7 51 unverified unverified JJ 12483 7 52 digital digital JJ 12483 7 53 content content NN 12483 7 54 will will MD 12483 7 55 be be VB 12483 7 56 served serve VBN 12483 7 57 by by IN 12483 7 58 world world NN 12483 7 59 ’s ’s POS 12483 7 60 libraries library NNS 12483 7 61 . . . 12483 8 1 The the DT 12483 8 2 solution solution NN 12483 8 3 is be VBZ 12483 8 4 to to TO 12483 8 5 develop develop VB 12483 8 6 automated automate VBN 12483 8 7 mechanisms mechanism NNS 12483 8 8 that that WDT 12483 8 9 can can MD 12483 8 10 check check VB 12483 8 11 the the DT 12483 8 12 credibility credibility NN 12483 8 13 of of IN 12483 8 14 digital digital JJ 12483 8 15 content content NN 12483 8 16 served serve VBD 12483 8 17 in in IN 12483 8 18 libraries library NNS 12483 8 19 without without IN 12483 8 20 manual manual JJ 12483 8 21 validation validation NN 12483 8 22 . . . 12483 9 1 For for IN 12483 9 2 this this DT 12483 9 3 purpose purpose NN 12483 9 4 , , , 12483 9 5 we -PRON- PRP 12483 9 6 developed develop VBD 12483 9 7 an an DT 12483 9 8 automated automated JJ 12483 9 9 fake fake JJ 12483 9 10 news news NN 12483 9 11 detection detection NN 12483 9 12 system system NN 12483 9 13 based base VBN 12483 9 14 on on IN 12483 9 15 Turkish turkish JJ 12483 9 16 digital digital JJ 12483 9 17 news news NN 12483 9 18 content content NN 12483 9 19 . . . 12483 10 1 Our -PRON- PRP$ 12483 10 2 approach approach NN 12483 10 3 can can MD 12483 10 4 be be VB 12483 10 5 modified modify VBN 12483 10 6 for for IN 12483 10 7 any any DT 12483 10 8 other other JJ 12483 10 9 language language NN 12483 10 10 if if IN 12483 10 11 there there EX 12483 10 12 is be VBZ 12483 10 13 labelled label VBN 12483 10 14 training training NN 12483 10 15 material material NN 12483 10 16 . . . 12483 11 1 This this DT 12483 11 2 model model NN 12483 11 3 can can MD 12483 11 4 be be VB 12483 11 5 integrated integrate VBN 12483 11 6 into into IN 12483 11 7 libraries library NNS 12483 11 8 ’ ’ POS 12483 11 9 digital digital JJ 12483 11 10 systems system NNS 12483 11 11 to to IN 12483 11 12 label label NNP 12483 11 13 served serve VBD 12483 11 14 news news NN 12483 11 15 content content NN 12483 11 16 as as IN 12483 11 17 potentially potentially RB 12483 11 18 fake fake JJ 12483 11 19 whenever whenever WRB 12483 11 20 necessary necessary JJ 12483 11 21 , , , 12483 11 22 preventing prevent VBG 12483 11 23 uncontrolled uncontrolled JJ 12483 11 24 falsehood falsehood NN 12483 11 25 dissemination dissemination NN 12483 11 26 via via IN 12483 11 27 libraries library NNS 12483 11 28 . . . 12483 12 1 INTRODUCTION introduction NN 12483 12 2 Collins Collins NNP 12483 12 3 dictionary dictionary JJ 12483 12 4 which which WDT 12483 12 5 chose choose VBD 12483 12 6 the the DT 12483 12 7 term term NN 12483 12 8 “ " `` 12483 12 9 fake fake JJ 12483 12 10 news news NN 12483 12 11 ” " '' 12483 12 12 as as IN 12483 12 13 the the DT 12483 12 14 “ " `` 12483 12 15 Word Word NNP 12483 12 16 of of IN 12483 12 17 the the DT 12483 12 18 Year year NN 12483 12 19 2017 2017 CD 12483 12 20 , , , 12483 12 21 ” " '' 12483 12 22 describes describe VBZ 12483 12 23 news news NN 12483 12 24 as as IN 12483 12 25 the the DT 12483 12 26 actual actual JJ 12483 12 27 and and CC 12483 12 28 objective objective JJ 12483 12 29 presentation presentation NN 12483 12 30 of of IN 12483 12 31 a a DT 12483 12 32 current current JJ 12483 12 33 event event NN 12483 12 34 , , , 12483 12 35 information information NN 12483 12 36 , , , 12483 12 37 or or CC 12483 12 38 situation situation NN 12483 12 39 that that WDT 12483 12 40 is be VBZ 12483 12 41 published publish VBN 12483 12 42 in in IN 12483 12 43 newspapers newspaper NNS 12483 12 44 and and CC 12483 12 45 broadcast broadcast VBN 12483 12 46 on on IN 12483 12 47 radio radio NN 12483 12 48 , , , 12483 12 49 television television NN 12483 12 50 , , , 12483 12 51 or or CC 12483 12 52 online.2 online.2 CD 12483 12 53 We -PRON- PRP 12483 12 54 are be VBP 12483 12 55 in in IN 12483 12 56 an an DT 12483 12 57 era era NN 12483 12 58 where where WRB 12483 12 59 everything everything NN 12483 12 60 goes go VBZ 12483 12 61 online online RB 12483 12 62 , , , 12483 12 63 and and CC 12483 12 64 news news NN 12483 12 65 is be VBZ 12483 12 66 not not RB 12483 12 67 an an DT 12483 12 68 exception exception NN 12483 12 69 . . . 12483 13 1 Many many JJ 12483 13 2 people people NNS 12483 13 3 today today NN 12483 13 4 prefer prefer VBP 12483 13 5 to to TO 12483 13 6 read read VB 12483 13 7 their -PRON- PRP$ 12483 13 8 daily daily JJ 12483 13 9 news news NN 12483 13 10 online online RB 12483 13 11 , , , 12483 13 12 because because IN 12483 13 13 it -PRON- PRP 12483 13 14 is be VBZ 12483 13 15 a a DT 12483 13 16 cost cost NN 12483 13 17 - - HYPH 12483 13 18 effective effective JJ 12483 13 19 and and CC 12483 13 20 convenient convenient JJ 12483 13 21 way way NN 12483 13 22 to to TO 12483 13 23 remain remain VB 12483 13 24 up up IN 12483 13 25 to to IN 12483 13 26 date date NN 12483 13 27 . . . 12483 14 1 Although although IN 12483 14 2 this this DT 12483 14 3 convenience convenience NN 12483 14 4 has have VBZ 12483 14 5 lucrative lucrative JJ 12483 14 6 benefits benefit NNS 12483 14 7 for for IN 12483 14 8 society society NN 12483 14 9 , , , 12483 14 10 it -PRON- PRP 12483 14 11 can can MD 12483 14 12 also also RB 12483 14 13 have have VB 12483 14 14 harmful harmful JJ 12483 14 15 side side JJ 12483 14 16 effects effect NNS 12483 14 17 . . . 12483 15 1 Having have VBG 12483 15 2 access access NN 12483 15 3 to to IN 12483 15 4 news news NN 12483 15 5 from from IN 12483 15 6 multiple multiple JJ 12483 15 7 sources source NNS 12483 15 8 , , , 12483 15 9 anytime anytime RB 12483 15 10 , , , 12483 15 11 anywhere anywhere RB 12483 15 12 has have VBZ 12483 15 13 become become VBN 12483 15 14 an an DT 12483 15 15 irresistible irresistible JJ 12483 15 16 part part NN 12483 15 17 of of IN 12483 15 18 our -PRON- PRP$ 12483 15 19 daily daily JJ 12483 15 20 routines routine NNS 12483 15 21 . . . 12483 16 1 However however RB 12483 16 2 , , , 12483 16 3 some some DT 12483 16 4 of of IN 12483 16 5 these these DT 12483 16 6 sources source NNS 12483 16 7 may may MD 12483 16 8 provide provide VB 12483 16 9 unverified unverified JJ 12483 16 10 content content NN 12483 16 11 which which WDT 12483 16 12 can can MD 12483 16 13 easily easily RB 12483 16 14 be be VB 12483 16 15 delivered deliver VBN 12483 16 16 right right RB 12483 16 17 to to IN 12483 16 18 your -PRON- PRP$ 12483 16 19 mobile mobile JJ 12483 16 20 device device NN 12483 16 21 . . . 12483 17 1 Most most RBS 12483 17 2 importantly importantly RB 12483 17 3 , , , 12483 17 4 potential potential JJ 12483 17 5 fake fake JJ 12483 17 6 news news NN 12483 17 7 content content NN 12483 17 8 delivered deliver VBN 12483 17 9 by by IN 12483 17 10 these these DT 12483 17 11 sources source NNS 12483 17 12 may may MD 12483 17 13 mislead mislead VB 12483 17 14 society society NN 12483 17 15 and and CC 12483 17 16 cause cause VB 12483 17 17 social social JJ 12483 17 18 disturbances disturbance NNS 12483 17 19 such such JJ 12483 17 20 as as IN 12483 17 21 triggering trigger VBG 12483 17 22 violence violence NN 12483 17 23 against against IN 12483 17 24 ethnic ethnic JJ 12483 17 25 minorities minority NNS 12483 17 26 and and CC 12483 17 27 refugees refugee NNS 12483 17 28 , , , 12483 17 29 causing cause VBG 12483 17 30 unnecessary unnecessary JJ 12483 17 31 fear fear NN 12483 17 32 related relate VBN 12483 17 33 to to IN 12483 17 34 health health NN 12483 17 35 issues issue NNS 12483 17 36 , , , 12483 17 37 or or CC 12483 17 38 even even RB 12483 17 39 sometimes sometimes RB 12483 17 40 result result VB 12483 17 41 in in IN 12483 17 42 crisis crisis NN 12483 17 43 , , , 12483 17 44 devastating devastating JJ 12483 17 45 riots riot NNS 12483 17 46 and and CC 12483 17 47 strikes strike NNS 12483 17 48 . . . 12483 18 1 Not not RB 12483 18 2 having have VBG 12483 18 3 a a DT 12483 18 4 steady steady JJ 12483 18 5 definition definition NN 12483 18 6 compared compare VBN 12483 18 7 to to IN 12483 18 8 news news NN 12483 18 9 , , , 12483 18 10 fake fake JJ 12483 18 11 news news NN 12483 18 12 is be VBZ 12483 18 13 often often RB 12483 18 14 defined define VBN 12483 18 15 according accord VBG 12483 18 16 to to IN 12483 18 17 the the DT 12483 18 18 data datum NNS 12483 18 19 used use VBN 12483 18 20 or or CC 12483 18 21 the the DT 12483 18 22 limited limited JJ 12483 18 23 perspective perspective NN 12483 18 24 of of IN 12483 18 25 the the DT 12483 18 26 study study NN 12483 18 27 in in IN 12483 18 28 the the DT 12483 18 29 literature literature NN 12483 18 30 . . . 12483 19 1 For for IN 12483 19 2 example example NN 12483 19 3 ; ; : 12483 19 4 DiFranzo DiFranzo NNP 12483 19 5 and and CC 12483 19 6 Gloria- Gloria- NNP 12483 19 7 Garcia Garcia NNP 12483 19 8 defined define VBD 12483 19 9 the the DT 12483 19 10 fake fake JJ 12483 19 11 news news NN 12483 19 12 as as IN 12483 19 13 “ " `` 12483 19 14 false false JJ 12483 19 15 news news NN 12483 19 16 stories story NNS 12483 19 17 that that WDT 12483 19 18 are be VBP 12483 19 19 packaged package VBN 12483 19 20 and and CC 12483 19 21 published publish VBN 12483 19 22 as as IN 12483 19 23 if if IN 12483 19 24 they -PRON- PRP 12483 19 25 were be VBD 12483 19 26 genuine genuine JJ 12483 19 27 . . . 12483 19 28 ”3 ”3 '' 12483 19 29 On on IN 12483 19 30 the the DT 12483 19 31 other other JJ 12483 19 32 hand hand NN 12483 19 33 , , , 12483 19 34 Guess Guess NNP 12483 19 35 et et FW 12483 19 36 al al NNP 12483 19 37 . . . 12483 20 1 see see VB 12483 20 2 the the DT 12483 20 3 term term NN 12483 20 4 as as IN 12483 20 5 “ " `` 12483 20 6 a a DT 12483 20 7 new new JJ 12483 20 8 form form NN 12483 20 9 of of IN 12483 20 10 political political JJ 12483 20 11 misinformation misinformation NN 12483 20 12 ” " '' 12483 20 13 within within IN 12483 20 14 the the DT 12483 20 15 domain domain NN 12483 20 16 of of IN 12483 20 17 politics politic NNS 12483 20 18 , , , 12483 20 19 whereas whereas IN 12483 20 20 Mustafaraj Mustafaraj NNPS 12483 20 21 is be VBZ 12483 20 22 more more RBR 12483 20 23 direct direct JJ 12483 20 24 and and CC 12483 20 25 defines define VBZ 12483 20 26 it -PRON- PRP 12483 20 27 as as IN 12483 20 28 mailto:umertoglu@hacettepe.edu.tr mailto:umertoglu@hacettepe.edu.tr NNP 12483 20 29 mailto:bgenc@cs.hacettepe.edu.tr mailto:bgenc@cs.hacettepe.edu.tr NNP 12483 20 30 INFORMATION INFORMATION VBD 12483 20 31 TECHNOLOGY TECHNOLOGY NNP 12483 20 32 AND and CC 12483 20 33 LIBRARIES library NNS 12483 20 34 DECEMBER DECEMBER NNP 12483 20 35 2020 2020 CD 12483 20 36 AUTOMATED AUTOMATED NNP 12483 20 37 FAKE FAKE NNP 12483 20 38 NEWS NEWS NNP 12483 20 39 DETECTION detection VB 12483 20 40 IN in IN 12483 20 41 THE the DT 12483 20 42 AGE AGE NNP 12483 20 43 OF of IN 12483 20 44 DIGITAL DIGITAL NNP 12483 20 45 LIBRARIES library NNS 12483 20 46 | | NNP 12483 20 47 MERTOĞLU MERTOĞLU NNS 12483 20 48 AND and CC 12483 20 49 GENÇ genç NN 12483 20 50 2 2 CD 12483 20 51 “ " `` 12483 20 52 lies lie NNS 12483 20 53 presented present VBN 12483 20 54 as as IN 12483 20 55 news news NN 12483 20 56 . . . 12483 20 57 ”4 ”4 NNP 12483 20 58 A a DT 12483 20 59 comprehensive comprehensive JJ 12483 20 60 list list NN 12483 20 61 of of IN 12483 20 62 12 12 CD 12483 20 63 definitions definition NNS 12483 20 64 can can MD 12483 20 65 be be VB 12483 20 66 found find VBN 12483 20 67 in in IN 12483 20 68 Egelhofer Egelhofer NNP 12483 20 69 and and CC 12483 20 70 Lecheler.5 Lecheler.5 NNP 12483 20 71 In in IN 12483 20 72 simplified simplified JJ 12483 20 73 terms term NNS 12483 20 74 , , , 12483 20 75 news news NN 12483 20 76 which which WDT 12483 20 77 is be VBZ 12483 20 78 created create VBN 12483 20 79 to to TO 12483 20 80 deceive deceive VB 12483 20 81 or or CC 12483 20 82 mislead mislead VB 12483 20 83 readers reader NNS 12483 20 84 can can MD 12483 20 85 be be VB 12483 20 86 called call VBN 12483 20 87 fake fake JJ 12483 20 88 news news NN 12483 20 89 . . . 12483 21 1 However however RB 12483 21 2 , , , 12483 21 3 the the DT 12483 21 4 concept concept NN 12483 21 5 of of IN 12483 21 6 fake fake JJ 12483 21 7 news news NN 12483 21 8 is be VBZ 12483 21 9 a a DT 12483 21 10 quite quite RB 12483 21 11 broad broad JJ 12483 21 12 one one NN 12483 21 13 that that WDT 12483 21 14 needs need VBZ 12483 21 15 to to TO 12483 21 16 be be VB 12483 21 17 specified specify VBN 12483 21 18 meticulously meticulously RB 12483 21 19 . . . 12483 22 1 Fake fake JJ 12483 22 2 news news NN 12483 22 3 is be VBZ 12483 22 4 created create VBN 12483 22 5 for for IN 12483 22 6 many many JJ 12483 22 7 purposes purpose NNS 12483 22 8 and and CC 12483 22 9 emerges emerge VBZ 12483 22 10 in in IN 12483 22 11 many many JJ 12483 22 12 different different JJ 12483 22 13 types type NNS 12483 22 14 . . . 12483 23 1 Having have VBG 12483 23 2 an an DT 12483 23 3 interwoven interwoven JJ 12483 23 4 structure structure NN 12483 23 5 , , , 12483 23 6 most most JJS 12483 23 7 of of IN 12483 23 8 these these DT 12483 23 9 types type NNS 12483 23 10 are be VBP 12483 23 11 shown show VBN 12483 23 12 in in IN 12483 23 13 figure figure NN 12483 23 14 1 1 CD 12483 23 15 . . . 12483 24 1 Although although IN 12483 24 2 , , , 12483 24 3 it -PRON- PRP 12483 24 4 is be VBZ 12483 24 5 not not RB 12483 24 6 easy easy JJ 12483 24 7 to to TO 12483 24 8 cluster cluster VB 12483 24 9 these these DT 12483 24 10 types type NNS 12483 24 11 into into IN 12483 24 12 separate separate JJ 12483 24 13 groups group NNS 12483 24 14 , , , 12483 24 15 they -PRON- PRP 12483 24 16 can can MD 12483 24 17 be be VB 12483 24 18 categorized categorize VBN 12483 24 19 according accord VBG 12483 24 20 to to IN 12483 24 21 the the DT 12483 24 22 information information NN 12483 24 23 quality quality NN 12483 24 24 or or CC 12483 24 25 based base VBN 12483 24 26 on on IN 12483 24 27 the the DT 12483 24 28 intention intention NN 12483 24 29 as as IN 12483 24 30 it -PRON- PRP 12483 24 31 is be VBZ 12483 24 32 created create VBN 12483 24 33 to to TO 12483 24 34 deceive deceive VB 12483 24 35 deliberately deliberately RB 12483 24 36 or or CC 12483 24 37 not not RB 12483 24 38 , , , 12483 24 39 as as IN 12483 24 40 Rashkin Rashkin NNP 12483 24 41 et et FW 12483 24 42 al al NNP 12483 24 43 . . . 12483 25 1 did.6 did.6 NFP 12483 25 2 We -PRON- PRP 12483 25 3 propose propose VBP 12483 25 4 the the DT 12483 25 5 following follow VBG 12483 25 6 classification classification NN 12483 25 7 where where WRB 12483 25 8 the the DT 12483 25 9 two two CD 12483 25 10 dimensions dimension NNS 12483 25 11 represent represent VBP 12483 25 12 the the DT 12483 25 13 potential potential JJ 12483 25 14 impact impact NN 12483 25 15 and and CC 12483 25 16 the the DT 12483 25 17 speed speed NN 12483 25 18 of of IN 12483 25 19 propagation propagation NN 12483 25 20 . . . 12483 26 1 Figure figure NN 12483 26 2 1 1 CD 12483 26 3 . . . 12483 27 1 The the DT 12483 27 2 volatile volatile JJ 12483 27 3 distribution distribution NN 12483 27 4 of of IN 12483 27 5 the the DT 12483 27 6 fake fake JJ 12483 27 7 news news NN 12483 27 8 types type NNS 12483 27 9 ( ( -LRB- 12483 27 10 clustered cluster VBN 12483 27 11 in in IN 12483 27 12 four four CD 12483 27 13 regions region NNS 12483 27 14 : : : 12483 27 15 sr sr NNP 12483 27 16 , , , 12483 27 17 Sr Sr NNP 12483 27 18 , , , 12483 27 19 Sr Sr NNP 12483 27 20 , , , 12483 27 21 SR SR NNP 12483 27 22 ) ) -RRB- 12483 27 23 with with IN 12483 27 24 respect respect NN 12483 27 25 to to IN 12483 27 26 two two CD 12483 27 27 dimensions dimension NNS 12483 27 28 : : : 12483 27 29 speed speed NN 12483 27 30 of of IN 12483 27 31 propagation propagation NN 12483 27 32 and and CC 12483 27 33 potential potential JJ 12483 27 34 impact impact NN 12483 27 35 . . . 12483 28 1 The the DT 12483 28 2 four four CD 12483 28 3 regions region NNS 12483 28 4 visualized visualize VBN 12483 28 5 are be VBP 12483 28 6 clustered cluster VBN 12483 28 7 according accord VBG 12483 28 8 to to IN 12483 28 9 their -PRON- PRP$ 12483 28 10 dangerousness dangerousness NN 12483 28 11 . . . 12483 29 1 First first RB 12483 29 2 of of IN 12483 29 3 all all DT 12483 29 4 , , , 12483 29 5 it -PRON- PRP 12483 29 6 should should MD 12483 29 7 be be VB 12483 29 8 noted note VBN 12483 29 9 that that IN 12483 29 10 to to TO 12483 29 11 order order VB 12483 29 12 types type NNS 12483 29 13 of of IN 12483 29 14 fake fake JJ 12483 29 15 news news NN 12483 29 16 in in IN 12483 29 17 a a DT 12483 29 18 stable stable JJ 12483 29 19 precision precision NN 12483 29 20 is be VBZ 12483 29 21 quite quite PDT 12483 29 22 a a DT 12483 29 23 challenging challenging JJ 12483 29 24 task task NN 12483 29 25 . . . 12483 30 1 The the DT 12483 30 2 variations variation NNS 12483 30 3 within within IN 12483 30 4 the the DT 12483 30 5 field field NN 12483 30 6 highly highly RB 12483 30 7 depend depend VBP 12483 30 8 on on IN 12483 30 9 dynamic dynamic JJ 12483 30 10 factors factor NNS 12483 30 11 such such JJ 12483 30 12 as as IN 12483 30 13 timespan timespan NN 12483 30 14 , , , 12483 30 15 actors actor NNS 12483 30 16 , , , 12483 30 17 and and CC 12483 30 18 echo- echo- NNP 12483 30 19 chamber chamber NN 12483 30 20 effect effect NN 12483 30 21 . . . 12483 31 1 Hence hence RB 12483 31 2 , , , 12483 31 3 this this DT 12483 31 4 figure figure NN 12483 31 5 should should MD 12483 31 6 be be VB 12483 31 7 considered consider VBN 12483 31 8 as as IN 12483 31 9 a a DT 12483 31 10 clustering cluster VBG 12483 31 11 effort effort NN 12483 31 12 . . . 12483 32 1 There there EX 12483 32 2 are be VBP 12483 32 3 possible possible JJ 12483 32 4 intersecting intersect VBG 12483 32 5 areas area NNS 12483 32 6 of of IN 12483 32 7 types type NNS 12483 32 8 within within IN 12483 32 9 the the DT 12483 32 10 regions region NNS 12483 32 11 . . . 12483 33 1 We -PRON- PRP 12483 33 2 will will MD 12483 33 3 now now RB 12483 33 4 give give VB 12483 33 5 examples example NNS 12483 33 6 for for IN 12483 33 7 two two CD 12483 33 8 regions region NNS 12483 33 9 , , , 12483 33 10 “ " `` 12483 33 11 sr sr NNP 12483 33 12 ” " '' 12483 33 13 and and CC 12483 33 14 “ " `` 12483 33 15 SR SR NNP 12483 33 16 . . . 12483 33 17 ” " '' 12483 33 18 For for IN 12483 33 19 example example NN 12483 33 20 , , , 12483 33 21 the the DT 12483 33 22 SR SR NNP 12483 33 23 grouping grouping NN 12483 33 24 shows show VBZ 12483 33 25 characteristics characteristic NNS 12483 33 26 of of IN 12483 33 27 high high JJ 12483 33 28 - - HYPH 12483 33 29 risk risk NN 12483 33 30 levels level NNS 12483 33 31 and and CC 12483 33 32 fast fast JJ 12483 33 33 dissemination dissemination NN 12483 33 34 . . . 12483 34 1 This this DT 12483 34 2 includes include VBZ 12483 34 3 varieties variety NNS 12483 34 4 of of IN 12483 34 5 fake fake JJ 12483 34 6 news news NN 12483 34 7 such such JJ 12483 34 8 as as IN 12483 34 9 propaganda propaganda NN 12483 34 10 , , , 12483 34 11 manipulation manipulation NN 12483 34 12 , , , 12483 34 13 misinformation misinformation NN 12483 34 14 , , , 12483 34 15 hate hate NN 12483 34 16 news news NN 12483 34 17 , , , 12483 34 18 INFORMATION INFORMATION NNP 12483 34 19 TECHNOLOGY TECHNOLOGY NNP 12483 34 20 AND and CC 12483 34 21 LIBRARIES library NNS 12483 34 22 DECEMBER DECEMBER NNP 12483 34 23 2020 2020 CD 12483 34 24 AUTOMATED AUTOMATED NNP 12483 34 25 FAKE FAKE NNP 12483 34 26 NEWS NEWS NNP 12483 34 27 DETECTION detection VB 12483 34 28 IN in IN 12483 34 29 THE the DT 12483 34 30 AGE AGE NNP 12483 34 31 OF of IN 12483 34 32 DIGITAL DIGITAL NNP 12483 34 33 LIBRARIES library NNS 12483 34 34 | | NNP 12483 34 35 MERTOĞLU MERTOĞLU NNS 12483 34 36 AND and CC 12483 34 37 GENÇ genç VBP 12483 34 38 3 3 CD 12483 34 39 provocative provocative JJ 12483 34 40 news news NN 12483 34 41 , , , 12483 34 42 etc etc FW 12483 34 43 . . . 12483 35 1 We -PRON- PRP 12483 35 2 usually usually RB 12483 35 3 encounter encounter VBP 12483 35 4 this this DT 12483 35 5 in in IN 12483 35 6 the the DT 12483 35 7 domain domain NN 12483 35 8 of of IN 12483 35 9 politics politic NNS 12483 35 10 . . . 12483 36 1 This this DT 12483 36 2 kind kind NN 12483 36 3 of of IN 12483 36 4 news news NN 12483 36 5 may may MD 12483 36 6 cause cause VB 12483 36 7 critical critical JJ 12483 36 8 and and CC 12483 36 9 nonrecoverable nonrecoverable JJ 12483 36 10 results result NNS 12483 36 11 in in IN 12483 36 12 politics politic NNS 12483 36 13 , , , 12483 36 14 the the DT 12483 36 15 economy economy NN 12483 36 16 , , , 12483 36 17 etc etc FW 12483 36 18 . . FW 12483 36 19 , , , 12483 36 20 in in IN 12483 36 21 a a DT 12483 36 22 short short JJ 12483 36 23 period period NN 12483 36 24 of of IN 12483 36 25 time time NN 12483 36 26 . . . 12483 37 1 The the DT 12483 37 2 rise rise NN 12483 37 3 of of IN 12483 37 4 the the DT 12483 37 5 term term NN 12483 37 6 fake fake JJ 12483 37 7 news news NN 12483 37 8 itself -PRON- PRP 12483 37 9 can can MD 12483 37 10 also also RB 12483 37 11 be be VB 12483 37 12 attributed attribute VBN 12483 37 13 to to IN 12483 37 14 this this DT 12483 37 15 kind kind NN 12483 37 16 of of IN 12483 37 17 news news NN 12483 37 18 . . . 12483 38 1 On on IN 12483 38 2 the the DT 12483 38 3 other other JJ 12483 38 4 hand hand NN 12483 38 5 , , , 12483 38 6 the the DT 12483 38 7 relatively relatively RB 12483 38 8 less less RBR 12483 38 9 severe severe JJ 12483 38 10 group group NN 12483 38 11 ( ( -LRB- 12483 38 12 sr sr NNP 12483 38 13 ) ) -RRB- 12483 38 14 of of IN 12483 38 15 fake fake JJ 12483 38 16 news news NN 12483 38 17 , , , 12483 38 18 comprising comprise VBG 12483 38 19 of of IN 12483 38 20 satire satire NN 12483 38 21 , , , 12483 38 22 hoax hoax NNP 12483 38 23 , , , 12483 38 24 click click NNP 12483 38 25 - - HYPH 12483 38 26 bait bait NN 12483 38 27 , , , 12483 38 28 etc etc FW 12483 38 29 . . . 12483 38 30 , , , 12483 38 31 has have VBZ 12483 38 32 low low JJ 12483 38 33 - - HYPH 12483 38 34 risk risk NN 12483 38 35 levels level NNS 12483 38 36 and and CC 12483 38 37 a a DT 12483 38 38 slow slow JJ 12483 38 39 speed speed NN 12483 38 40 of of IN 12483 38 41 dissemination dissemination NN 12483 38 42 . . . 12483 39 1 A a DT 12483 39 2 frequently frequently RB 12483 39 3 used use VBN 12483 39 4 type type NN 12483 39 5 of of IN 12483 39 6 this this DT 12483 39 7 group group NN 12483 39 8 , , , 12483 39 9 click click NN 12483 39 10 - - HYPH 12483 39 11 bait bait NN 12483 39 12 , , , 12483 39 13 is be VBZ 12483 39 14 a a DT 12483 39 15 sensational sensational JJ 12483 39 16 headline headline NN 12483 39 17 or or CC 12483 39 18 link link NN 12483 39 19 that that WDT 12483 39 20 urges urge VBZ 12483 39 21 the the DT 12483 39 22 reader reader NN 12483 39 23 to to TO 12483 39 24 click click VB 12483 39 25 on on IN 12483 39 26 a a DT 12483 39 27 post post NN 12483 39 28 , , , 12483 39 29 link link NN 12483 39 30 , , , 12483 39 31 article article NN 12483 39 32 , , , 12483 39 33 image image NN 12483 39 34 , , , 12483 39 35 or or CC 12483 39 36 video video NN 12483 39 37 . . . 12483 40 1 These these DT 12483 40 2 kinds kind NNS 12483 40 3 of of IN 12483 40 4 news news NN 12483 40 5 have have VBP 12483 40 6 a a DT 12483 40 7 repetitive repetitive JJ 12483 40 8 style style NN 12483 40 9 . . . 12483 41 1 It -PRON- PRP 12483 41 2 can can MD 12483 41 3 be be VB 12483 41 4 said say VBN 12483 41 5 that that IN 12483 41 6 readers reader NNS 12483 41 7 become become VBP 12483 41 8 aware aware JJ 12483 41 9 of of IN 12483 41 10 falsehood falsehood NN 12483 41 11 after after IN 12483 41 12 experiencing experience VBG 12483 41 13 a a DT 12483 41 14 few few JJ 12483 41 15 times time NNS 12483 41 16 . . . 12483 42 1 So so RB 12483 42 2 , , , 12483 42 3 risk risk NN 12483 42 4 level level NN 12483 42 5 is be VBZ 12483 42 6 lower low JJR 12483 42 7 , , , 12483 42 8 and and CC 12483 42 9 dissemination dissemination NN 12483 42 10 is be VBZ 12483 42 11 slower slow JJR 12483 42 12 . . . 12483 43 1 Vosoughi Vosoughi NNP 12483 43 2 et et NNP 12483 43 3 al al NNP 12483 43 4 . . . 12483 44 1 stated state VBD 12483 44 2 the the DT 12483 44 3 assumption assumption NN 12483 44 4 that that IN 12483 44 5 “ " `` 12483 44 6 Falsehood falsehood NN 12483 44 7 diffuses diffuse NNS 12483 44 8 significantly significantly RB 12483 44 9 farther farther RB 12483 44 10 , , , 12483 44 11 faster fast RBR 12483 44 12 , , , 12483 44 13 deeper deep JJR 12483 44 14 , , , 12483 44 15 and and CC 12483 44 16 more more RBR 12483 44 17 broadly broadly RB 12483 44 18 than than IN 12483 44 19 the the DT 12483 44 20 truth truth NN 12483 44 21 . . . 12483 44 22 ”7 ”7 '' 12483 44 23 So so RB 12483 44 24 indeed indeed RB 12483 44 25 , , , 12483 44 26 just just RB 12483 44 27 one one CD 12483 44 28 piece piece NN 12483 44 29 of of IN 12483 44 30 fake fake JJ 12483 44 31 news news NN 12483 44 32 may may MD 12483 44 33 affect affect VB 12483 44 34 many many JJ 12483 44 35 more more JJR 12483 44 36 people people NNS 12483 44 37 than than IN 12483 44 38 thousands thousand NNS 12483 44 39 of of IN 12483 44 40 true true JJ 12483 44 41 news news NN 12483 44 42 items item NNS 12483 44 43 do do VBP 12483 44 44 because because IN 12483 44 45 of of IN 12483 44 46 the the DT 12483 44 47 dramatic dramatic JJ 12483 44 48 circulation circulation NN 12483 44 49 of of IN 12483 44 50 fake fake JJ 12483 44 51 news news NN 12483 44 52 . . . 12483 45 1 In in IN 12483 45 2 their -PRON- PRP$ 12483 45 3 recent recent JJ 12483 45 4 survey survey NN 12483 45 5 about about IN 12483 45 6 fake fake JJ 12483 45 7 news news NN 12483 45 8 , , , 12483 45 9 Zhou Zhou NNP 12483 45 10 and and CC 12483 45 11 Zafarani Zafarani NNP 12483 45 12 highlighted highlight VBD 12483 45 13 that that IN 12483 45 14 fake fake JJ 12483 45 15 news news NN 12483 45 16 is be VBZ 12483 45 17 a a DT 12483 45 18 major major JJ 12483 45 19 concern concern NN 12483 45 20 for for IN 12483 45 21 many many JJ 12483 45 22 different different JJ 12483 45 23 research research NN 12483 45 24 disciplines discipline NNS 12483 45 25 especially especially RB 12483 45 26 information information NN 12483 45 27 technologies technology NNS 12483 45 28 . . . 12483 46 1 8 8 LS 12483 46 2 Being be VBG 12483 46 3 a a DT 12483 46 4 trusted trust VBN 12483 46 5 source source NN 12483 46 6 of of IN 12483 46 7 information information NN 12483 46 8 for for IN 12483 46 9 a a DT 12483 46 10 long long JJ 12483 46 11 time time NN 12483 46 12 , , , 12483 46 13 libraries library NNS 12483 46 14 will will MD 12483 46 15 play play VB 12483 46 16 an an DT 12483 46 17 important important JJ 12483 46 18 role role NN 12483 46 19 in in IN 12483 46 20 fighting fighting NN 12483 46 21 against against IN 12483 46 22 fake fake JJ 12483 46 23 news news NN 12483 46 24 problem problem NN 12483 46 25 . . . 12483 47 1 Kattimani Kattimani NNP 12483 47 2 et et FW 12483 47 3 al al NNP 12483 47 4 . . . 12483 48 1 claims claim NNS 12483 48 2 that that IN 12483 48 3 the the DT 12483 48 4 modern modern JJ 12483 48 5 librarian librarian NN 12483 48 6 must must MD 12483 48 7 be be VB 12483 48 8 equipped equip VBN 12483 48 9 with with IN 12483 48 10 necessary necessary JJ 12483 48 11 digital digital JJ 12483 48 12 skills skill NNS 12483 48 13 and and CC 12483 48 14 tools tool NNS 12483 48 15 to to TO 12483 48 16 handle handle VB 12483 48 17 both both DT 12483 48 18 printed print VBN 12483 48 19 collections collection NNS 12483 48 20 and and CC 12483 48 21 newly newly RB 12483 48 22 emerging emerge VBG 12483 48 23 digital digital NNP 12483 48 24 resources.9 resources.9 NNP 12483 48 25 Similarly similarly RB 12483 48 26 , , , 12483 48 27 we -PRON- PRP 12483 48 28 foresee foresee VBP 12483 48 29 that that IN 12483 48 30 digital digital JJ 12483 48 31 libraries library NNS 12483 48 32 , , , 12483 48 33 which which WDT 12483 48 34 can can MD 12483 48 35 be be VB 12483 48 36 defined define VBN 12483 48 37 as as IN 12483 48 38 collections collection NNS 12483 48 39 of of IN 12483 48 40 digital digital JJ 12483 48 41 content content NN 12483 48 42 licensed license VBN 12483 48 43 and and CC 12483 48 44 maintained maintain VBN 12483 48 45 by by IN 12483 48 46 libraries library NNS 12483 48 47 , , , 12483 48 48 can can MD 12483 48 49 be be VB 12483 48 50 a a DT 12483 48 51 part part NN 12483 48 52 of of IN 12483 48 53 the the DT 12483 48 54 solution solution NN 12483 48 55 as as IN 12483 48 56 an an DT 12483 48 57 authority authority NN 12483 48 58 service service NN 12483 48 59 with with IN 12483 48 60 a a DT 12483 48 61 collective collective JJ 12483 48 62 effort effort NN 12483 48 63 . . . 12483 49 1 Connaway Connaway NNP 12483 49 2 et et FW 12483 49 3 al al NNP 12483 49 4 . . . 12483 50 1 point point NN 12483 50 2 to to IN 12483 50 3 the the DT 12483 50 4 key key JJ 12483 50 5 role role NN 12483 50 6 of of IN 12483 50 7 information information NN 12483 50 8 professionals professional NNS 12483 50 9 such such JJ 12483 50 10 as as IN 12483 50 11 librarians librarian NNS 12483 50 12 , , , 12483 50 13 archivists archivist NNS 12483 50 14 , , , 12483 50 15 journalists journalist NNS 12483 50 16 , , , 12483 50 17 and and CC 12483 50 18 information information NN 12483 50 19 architects architect NNS 12483 50 20 in in IN 12483 50 21 helping help VBG 12483 50 22 society society NN 12483 50 23 use use VB 12483 50 24 the the DT 12483 50 25 products product NNS 12483 50 26 and and CC 12483 50 27 services service NNS 12483 50 28 related relate VBN 12483 50 29 to to IN 12483 50 30 news news NN 12483 50 31 in in IN 12483 50 32 a a DT 12483 50 33 convenient convenient JJ 12483 50 34 way way NN 12483 50 35 . . . 12483 51 1 10 10 CD 12483 51 2 As as IN 12483 51 3 libraries library NNS 12483 51 4 all all RB 12483 51 5 over over IN 12483 51 6 the the DT 12483 51 7 world world NN 12483 51 8 are be VBP 12483 51 9 transitioning transition VBG 12483 51 10 into into IN 12483 51 11 digital digital JJ 12483 51 12 content content NN 12483 51 13 delivery delivery NN 12483 51 14 services service NNS 12483 51 15 , , , 12483 51 16 they -PRON- PRP 12483 51 17 should should MD 12483 51 18 implement implement VB 12483 51 19 mechanisms mechanism NNS 12483 51 20 to to TO 12483 51 21 avoid avoid VB 12483 51 22 fake fake JJ 12483 51 23 and and CC 12483 51 24 misleading misleading JJ 12483 51 25 content content NN 12483 51 26 being be VBG 12483 51 27 disseminated disseminate VBN 12483 51 28 through through IN 12483 51 29 them -PRON- PRP 12483 51 30 under under IN 12483 51 31 the the DT 12483 51 32 guidance guidance NN 12483 51 33 of of IN 12483 51 34 information information NN 12483 51 35 professionals professional NNS 12483 51 36 . . . 12483 52 1 To to TO 12483 52 2 lay lay VB 12483 52 3 out out RP 12483 52 4 proper proper JJ 12483 52 5 future future JJ 12483 52 6 directions direction NNS 12483 52 7 for for IN 12483 52 8 the the DT 12483 52 9 solution solution NN 12483 52 10 strategy strategy NN 12483 52 11 , , , 12483 52 12 a a DT 12483 52 13 clear clear JJ 12483 52 14 understanding understanding NN 12483 52 15 of of IN 12483 52 16 interaction interaction NN 12483 52 17 between between IN 12483 52 18 library library NN 12483 52 19 and and CC 12483 52 20 information information NN 12483 52 21 science science NN 12483 52 22 ( ( -LRB- 12483 52 23 LIS LIS NNP 12483 52 24 ) ) -RRB- 12483 52 25 community community NN 12483 52 26 and and CC 12483 52 27 fake fake JJ 12483 52 28 news news NN 12483 52 29 must must MD 12483 52 30 be be VB 12483 52 31 addressed address VBN 12483 52 32 . . . 12483 53 1 Sullivan Sullivan NNP 12483 53 2 states state VBZ 12483 53 3 that that IN 12483 53 4 the the DT 12483 53 5 LIS LIS NNP 12483 53 6 community community NN 12483 53 7 has have VBZ 12483 53 8 been be VBN 12483 53 9 affected affect VBN 12483 53 10 deeply deeply RB 12483 53 11 in in IN 12483 53 12 the the DT 12483 53 13 aftermath aftermath NN 12483 53 14 of of IN 12483 53 15 the the DT 12483 53 16 2016 2016 CD 12483 53 17 US US NNP 12483 53 18 presidential presidential JJ 12483 53 19 elections.11 elections.11 NNP 12483 53 20 Moreover moreover RB 12483 53 21 , , , 12483 53 22 he -PRON- PRP 12483 53 23 quotes quote VBZ 12483 53 24 many many JJ 12483 53 25 other other JJ 12483 53 26 scientists scientist NNS 12483 53 27 , , , 12483 53 28 emphasizing emphasize VBG 12483 53 29 libraries library NNS 12483 53 30 ’ ' '' 12483 53 31 and and CC 12483 53 32 librarians librarian NNS 12483 53 33 ’ ’ POS 12483 53 34 role role NN 12483 53 35 in in IN 12483 53 36 the the DT 12483 53 37 fight fight NN 12483 53 38 against against IN 12483 53 39 fake fake JJ 12483 53 40 news news NN 12483 53 41 . . . 12483 54 1 For for IN 12483 54 2 example example NN 12483 54 3 , , , 12483 54 4 Finley Finley NNP 12483 54 5 et et NNP 12483 54 6 al al NNP 12483 54 7 . . . 12483 55 1 say say VBP 12483 55 2 that that IN 12483 55 3 libraries library NNS 12483 55 4 are be VBP 12483 55 5 the the DT 12483 55 6 direct direct JJ 12483 55 7 antithesis antithesis NN 12483 55 8 of of IN 12483 55 9 fake fake JJ 12483 55 10 news news NN 12483 55 11 , , , 12483 55 12 the the DT 12483 55 13 American American NNP 12483 55 14 Library Library NNP 12483 55 15 Association Association NNP 12483 55 16 ( ( -LRB- 12483 55 17 ALA ALA NNP 12483 55 18 ) ) -RRB- 12483 55 19 called call VBD 12483 55 20 fake fake JJ 12483 55 21 news news NN 12483 55 22 an an DT 12483 55 23 anathema anathema NN 12483 55 24 to to IN 12483 55 25 the the DT 12483 55 26 ethics ethic NNS 12483 55 27 of of IN 12483 55 28 librarianship librarianship NN 12483 55 29 in in IN 12483 55 30 2017 2017 CD 12483 55 31 , , , 12483 55 32 Rochlin Rochlin NNP 12483 55 33 emphasizes emphasize VBZ 12483 55 34 the the DT 12483 55 35 role role NN 12483 55 36 of of IN 12483 55 37 librarians librarian NNS 12483 55 38 in in IN 12483 55 39 this this DT 12483 55 40 fight fight NN 12483 55 41 , , , 12483 55 42 and and CC 12483 55 43 talks talk NNS 12483 55 44 about about IN 12483 55 45 the the DT 12483 55 46 need need NN 12483 55 47 to to TO 12483 55 48 adopt adopt VB 12483 55 49 fake fake JJ 12483 55 50 news news NN 12483 55 51 as as IN 12483 55 52 a a DT 12483 55 53 central central JJ 12483 55 54 concern concern NN 12483 55 55 in in IN 12483 55 56 librarianship librarianship NN 12483 55 57 and and CC 12483 55 58 many many JJ 12483 55 59 other other JJ 12483 55 60 researchers researcher NNS 12483 55 61 name name VBP 12483 55 62 librarians librarian NNS 12483 55 63 in in IN 12483 55 64 the the DT 12483 55 65 front front JJ 12483 55 66 lines line NNS 12483 55 67 of of IN 12483 55 68 the the DT 12483 55 69 fight fight NN 12483 55 70 against against IN 12483 55 71 fake fake JJ 12483 55 72 news.12 news.12 NNP 12483 55 73 Today Today NNP 12483 55 74 , , , 12483 55 75 the the DT 12483 55 76 struggle struggle NN 12483 55 77 to to TO 12483 55 78 detect detect VB 12483 55 79 fake fake JJ 12483 55 80 news news NN 12483 55 81 and and CC 12483 55 82 prevent prevent VB 12483 55 83 their -PRON- PRP$ 12483 55 84 spread spread NN 12483 55 85 is be VBZ 12483 55 86 so so RB 12483 55 87 popular popular JJ 12483 55 88 that that IN 12483 55 89 competitions competition NNS 12483 55 90 are be VBP 12483 55 91 being be VBG 12483 55 92 organized organize VBN 12483 55 93 ( ( -LRB- 12483 55 94 e.g. e.g. RB 12483 55 95 , , , 12483 55 96 http://www.fakenewschallenge.org/ http://www.fakenewschallenge.org/ ADD 12483 55 97 ) ) -RRB- 12483 55 98 and and CC 12483 55 99 conferences conference NNS 12483 55 100 are be VBP 12483 55 101 being be VBG 12483 55 102 held hold VBN 12483 55 103 ( ( -LRB- 12483 55 104 e.g. e.g. RB 12483 55 105 , , , 12483 55 106 Bobcatsss Bobcatsss NNP 12483 55 107 2020 2020 CD 12483 55 108 ) ) -RRB- 12483 55 109 . . . 12483 56 1 The the DT 12483 56 2 struggle struggle NN 12483 56 3 against against IN 12483 56 4 fake fake JJ 12483 56 5 news news NN 12483 56 6 can can MD 12483 56 7 be be VB 12483 56 8 classified classify VBN 12483 56 9 under under IN 12483 56 10 three three CD 12483 56 11 main main JJ 12483 56 12 venues venue NNS 12483 56 13 : : : 12483 56 14 • • VB 12483 56 15 Reader reader NN 12483 56 16 awareness awareness NN 12483 56 17 • • VBP 12483 56 18 Fact fact NN 12483 56 19 - - HYPH 12483 56 20 checking check VBG 12483 56 21 organizations organization NNS 12483 56 22 and and CC 12483 56 23 websites website NNS 12483 56 24 • • VBP 12483 56 25 Automated automate VBN 12483 56 26 detection detection NN 12483 56 27 systems system VBZ 12483 56 28 The the DT 12483 56 29 first first JJ 12483 56 30 item item NN 12483 56 31 requires require VBZ 12483 56 32 awareness awareness NN 12483 56 33 of of IN 12483 56 34 individuals individual NNS 12483 56 35 against against IN 12483 56 36 fake fake JJ 12483 56 37 news news NN 12483 56 38 and and CC 12483 56 39 a a DT 12483 56 40 collective collective JJ 12483 56 41 conscience conscience NN 12483 56 42 within within IN 12483 56 43 the the DT 12483 56 44 society society NN 12483 56 45 against against IN 12483 56 46 spreading spread VBG 12483 56 47 fake fake JJ 12483 56 48 news news NN 12483 56 49 . . . 12483 57 1 To to IN 12483 57 2 this this DT 12483 57 3 end end NN 12483 57 4 , , , 12483 57 5 visual visual JJ 12483 57 6 and and CC 12483 57 7 textual textual JJ 12483 57 8 checklists checklist NNS 12483 57 9 , , , 12483 57 10 frameworks framework NNS 12483 57 11 , , , 12483 57 12 and and CC 12483 57 13 guidance guidance NN 12483 57 14 lists list NNS 12483 57 15 are be VBP 12483 57 16 being be VBG 12483 57 17 published publish VBN 12483 57 18 by by IN 12483 57 19 official official JJ 12483 57 20 organizations organization NNS 12483 57 21 , , , 12483 57 22 such such JJ 12483 57 23 as as IN 12483 57 24 IFLA’s13 ifla’s13 FW 12483 57 25 http://www.fakenewschallenge.org/ http://www.fakenewschallenge.org/ VB 12483 57 26 INFORMATION INFORMATION NNP 12483 57 27 TECHNOLOGY TECHNOLOGY NNP 12483 57 28 AND and CC 12483 57 29 LIBRARIES library NNS 12483 57 30 DECEMBER DECEMBER NNP 12483 57 31 2020 2020 CD 12483 57 32 AUTOMATED AUTOMATED NNP 12483 57 33 FAKE FAKE NNP 12483 57 34 NEWS NEWS NNP 12483 57 35 DETECTION detection VB 12483 57 36 IN in IN 12483 57 37 THE the DT 12483 57 38 AGE AGE NNP 12483 57 39 OF of IN 12483 57 40 DIGITAL DIGITAL NNP 12483 57 41 LIBRARIES library NNS 12483 57 42 | | NNP 12483 57 43 MERTOĞLU MERTOĞLU NNS 12483 57 44 AND and CC 12483 57 45 GENÇ genç NN 12483 57 46 4 4 CD 12483 57 47 ( ( -LRB- 12483 57 48 International International NNP 12483 57 49 Federation Federation NNP 12483 57 50 of of IN 12483 57 51 Library Library NNP 12483 57 52 Associations Associations NNPS 12483 57 53 ) ) -RRB- 12483 57 54 infographic infographic JJ 12483 57 55 which which WDT 12483 57 56 contains contain VBZ 12483 57 57 eight eight CD 12483 57 58 steps step NNS 12483 57 59 to to TO 12483 57 60 spot spot VB 12483 57 61 fake fake JJ 12483 57 62 news news NN 12483 57 63 . . . 12483 58 1 The the DT 12483 58 2 RADAR RADAR NNS 12483 58 3 framework framework NN 12483 58 4 and and CC 12483 58 5 the the DT 12483 58 6 Currency Currency NNP 12483 58 7 , , , 12483 58 8 Relevance Relevance NNP 12483 58 9 , , , 12483 58 10 Authority Authority NNP 12483 58 11 , , , 12483 58 12 Accuracy Accuracy NNP 12483 58 13 , , , 12483 58 14 and and CC 12483 58 15 Purpose Purpose NNP 12483 58 16 ( ( -LRB- 12483 58 17 CRAAP CRAAP NNP 12483 58 18 ) ) -RRB- 12483 58 19 test test NN 12483 58 20 are be VBP 12483 58 21 some some DT 12483 58 22 of of IN 12483 58 23 the the DT 12483 58 24 efforts effort NNS 12483 58 25 trying try VBG 12483 58 26 to to TO 12483 58 27 increase increase VB 12483 58 28 reader reader NN 12483 58 29 - - HYPH 12483 58 30 awareness awareness NN 12483 58 31 of of IN 12483 58 32 fake fake JJ 12483 58 33 news.14 news.14 NNP 12483 58 34 Unfortunately unfortunately RB 12483 58 35 , , , 12483 58 36 due due IN 12483 58 37 to to IN 12483 58 38 the the DT 12483 58 39 nature nature NN 12483 58 40 of of IN 12483 58 41 fake fake JJ 12483 58 42 news news NN 12483 58 43 and and CC 12483 58 44 the the DT 12483 58 45 clever clever JJ 12483 58 46 way way NN 12483 58 47 they -PRON- PRP 12483 58 48 are be VBP 12483 58 49 created create VBN 12483 58 50 triggering trigger VBG 12483 58 51 people people NNS 12483 58 52 ’s ’s POS 12483 58 53 hunger hunger NN 12483 58 54 to to TO 12483 58 55 spread spread VB 12483 58 56 sensational sensational JJ 12483 58 57 information information NN 12483 58 58 , , , 12483 58 59 it -PRON- PRP 12483 58 60 is be VBZ 12483 58 61 very very RB 12483 58 62 difficult difficult JJ 12483 58 63 to to TO 12483 58 64 achieve achieve VB 12483 58 65 full full JJ 12483 58 66 control control NN 12483 58 67 via via IN 12483 58 68 this this DT 12483 58 69 strategy strategy NN 12483 58 70 . . . 12483 59 1 Some some DT 12483 59 2 studies study NNS 12483 59 3 explicitly explicitly RB 12483 59 4 showed show VBD 12483 59 5 that that IN 12483 59 6 humans human NNS 12483 59 7 are be VBP 12483 59 8 prone prone JJ 12483 59 9 to to TO 12483 59 10 get get VB 12483 59 11 confused confused JJ 12483 59 12 when when WRB 12483 59 13 it -PRON- PRP 12483 59 14 comes come VBZ 12483 59 15 to to IN 12483 59 16 spotting spot VBG 12483 59 17 lies lie NNS 12483 59 18 or or CC 12483 59 19 deciding decide VBG 12483 59 20 whether whether IN 12483 59 21 a a DT 12483 59 22 news news NN 12483 59 23 item item NN 12483 59 24 is be VBZ 12483 59 25 fake fake JJ 12483 59 26 or or CC 12483 59 27 not.15 not.15 NNP 12483 59 28 Furthermore furthermore RB 12483 59 29 , , , 12483 59 30 people people NNS 12483 59 31 often often RB 12483 59 32 overlook overlook VBP 12483 59 33 facts fact NNS 12483 59 34 that that WDT 12483 59 35 conflict conflict NN 12483 59 36 with with IN 12483 59 37 their -PRON- PRP$ 12483 59 38 current current JJ 12483 59 39 belief belief NN 12483 59 40 , , , 12483 59 41 especially especially RB 12483 59 42 in in IN 12483 59 43 politics politic NNS 12483 59 44 and and CC 12483 59 45 controversial controversial JJ 12483 59 46 social social JJ 12483 59 47 issues.16 issues.16 NNP 12483 59 48 The the DT 12483 59 49 second second JJ 12483 59 50 strategy strategy NN 12483 59 51 focuses focus VBZ 12483 59 52 on on IN 12483 59 53 third third JJ 12483 59 54 - - HYPH 12483 59 55 party party NN 12483 59 56 manually manually RB 12483 59 57 driven drive VBN 12483 59 58 systems system NNS 12483 59 59 for for IN 12483 59 60 checking checking NN 12483 59 61 and and CC 12483 59 62 labelling labelling NN 12483 59 63 content content NN 12483 59 64 as as IN 12483 59 65 fake fake JJ 12483 59 66 or or CC 12483 59 67 valid valid JJ 12483 59 68 . . . 12483 60 1 Recently recently RB 12483 60 2 , , , 12483 60 3 we -PRON- PRP 12483 60 4 have have VBP 12483 60 5 seen see VBN 12483 60 6 many many JJ 12483 60 7 examples example NNS 12483 60 8 of of IN 12483 60 9 offline offline NN 12483 60 10 and and CC 12483 60 11 online online JJ 12483 60 12 organizations organization NNS 12483 60 13 trying try VBG 12483 60 14 to to TO 12483 60 15 work work VB 12483 60 16 according accord VBG 12483 60 17 to to IN 12483 60 18 this this DT 12483 60 19 strategy strategy NN 12483 60 20 , , , 12483 60 21 such such JJ 12483 60 22 as as IN 12483 60 23 a a DT 12483 60 24 growing grow VBG 12483 60 25 body body NN 12483 60 26 of of IN 12483 60 27 fact fact NN 12483 60 28 - - HYPH 12483 60 29 checking check VBG 12483 60 30 organizations organization NNS 12483 60 31 , , , 12483 60 32 start start NN 12483 60 33 - - HYPH 12483 60 34 ups up NNS 12483 60 35 ( ( -LRB- 12483 60 36 Storyzy Storyzy NNP 12483 60 37 , , , 12483 60 38 Factmata Factmata NNP 12483 60 39 , , , 12483 60 40 etc etc FW 12483 60 41 . . . 12483 61 1 ) ) -RRB- 12483 61 2 , , , 12483 61 3 and and CC 12483 61 4 other other JJ 12483 61 5 projects project NNS 12483 61 6 with with IN 12483 61 7 similar similar JJ 12483 61 8 purposes.17 purposes.17 NNP 12483 61 9 Unfortunately unfortunately RB 12483 61 10 , , , 12483 61 11 these these DT 12483 61 12 manually manually RB 12483 61 13 powered power VBN 12483 61 14 systems system NNS 12483 61 15 can can MD 12483 61 16 not not RB 12483 61 17 cope cope VB 12483 61 18 with with IN 12483 61 19 the the DT 12483 61 20 huge huge JJ 12483 61 21 amounts amount NNS 12483 61 22 of of IN 12483 61 23 digital digital JJ 12483 61 24 content content NN 12483 61 25 being be VBG 12483 61 26 steadily steadily RB 12483 61 27 produced produce VBN 12483 61 28 . . . 12483 62 1 Therefore therefore RB 12483 62 2 , , , 12483 62 3 they -PRON- PRP 12483 62 4 focus focus VBP 12483 62 5 only only RB 12483 62 6 on on IN 12483 62 7 a a DT 12483 62 8 subset subset NN 12483 62 9 of of IN 12483 62 10 digital digital JJ 12483 62 11 content content NN 12483 62 12 that that WDT 12483 62 13 they -PRON- PRP 12483 62 14 classify classify VBP 12483 62 15 as as IN 12483 62 16 having have VBG 12483 62 17 higher high JJR 12483 62 18 priority priority NN 12483 62 19 . . . 12483 63 1 Even even RB 12483 63 2 for for IN 12483 63 3 this this DT 12483 63 4 subset subset NN 12483 63 5 of of IN 12483 63 6 content content NN 12483 63 7 , , , 12483 63 8 their -PRON- PRP$ 12483 63 9 reaction reaction NN 12483 63 10 speed speed NN 12483 63 11 is be VBZ 12483 63 12 much much RB 12483 63 13 slower slow JJR 12483 63 14 than than IN 12483 63 15 the the DT 12483 63 16 fake fake JJ 12483 63 17 information information NN 12483 63 18 ’s ’s , 12483 63 19 spread spread VBD 12483 63 20 speed speed NN 12483 63 21 . . . 12483 64 1 Therefore therefore RB 12483 64 2 , , , 12483 64 3 automated automate VBN 12483 64 4 and and CC 12483 64 5 verified verify VBD 12483 64 6 systems system NNS 12483 64 7 emerge emerge VBP 12483 64 8 as as IN 12483 64 9 an an DT 12483 64 10 inevitable inevitable JJ 12483 64 11 last last JJ 12483 64 12 option option NN 12483 64 13 . . . 12483 65 1 The the DT 12483 65 2 third third JJ 12483 65 3 strategy strategy NN 12483 65 4 offers offer VBZ 12483 65 5 automated automate VBD 12483 65 6 fact fact NN 12483 65 7 - - HYPH 12483 65 8 checking check VBG 12483 65 9 systems system NNS 12483 65 10 , , , 12483 65 11 which which WDT 12483 65 12 once once RB 12483 65 13 trained train VBD 12483 65 14 , , , 12483 65 15 can can MD 12483 65 16 deliver deliver VB 12483 65 17 content content NN 12483 65 18 labelling labelling NN 12483 65 19 at at IN 12483 65 20 unprecedented unprecedented JJ 12483 65 21 speeds speed NNS 12483 65 22 . . . 12483 66 1 Today today NN 12483 66 2 , , , 12483 66 3 many many JJ 12483 66 4 researchers researcher NNS 12483 66 5 are be VBP 12483 66 6 researching research VBG 12483 66 7 automated automate VBN 12483 66 8 solutions solution NNS 12483 66 9 and and CC 12483 66 10 building building NN 12483 66 11 models model NNS 12483 66 12 with with IN 12483 66 13 different different JJ 12483 66 14 methodologies.18 methodologies.18 NNP 12483 66 15 Notwithstanding Notwithstanding NNP 12483 66 16 the the DT 12483 66 17 latest late JJS 12483 66 18 studies study NNS 12483 66 19 , , , 12483 66 20 there there EX 12483 66 21 is be VBZ 12483 66 22 still still RB 12483 66 23 a a DT 12483 66 24 lot lot NN 12483 66 25 to to TO 12483 66 26 do do VB 12483 66 27 in in IN 12483 66 28 the the DT 12483 66 29 realm realm NN 12483 66 30 of of IN 12483 66 31 automated automate VBN 12483 66 32 fake fake JJ 12483 66 33 news news NN 12483 66 34 detection detection NN 12483 66 35 . . . 12483 67 1 Automated automate VBN 12483 67 2 fact fact NN 12483 67 3 - - HYPH 12483 67 4 checking check VBG 12483 67 5 systems system NNS 12483 67 6 will will MD 12483 67 7 be be VB 12483 67 8 detailed detail VBN 12483 67 9 in in IN 12483 67 10 the the DT 12483 67 11 rest rest NN 12483 67 12 of of IN 12483 67 13 the the DT 12483 67 14 paper paper NN 12483 67 15 . . . 12483 68 1 Thanks thank NNS 12483 68 2 to to IN 12483 68 3 the the DT 12483 68 4 internet internet NN 12483 68 5 , , , 12483 68 6 the the DT 12483 68 7 collections collection NNS 12483 68 8 of of IN 12483 68 9 digital digital JJ 12483 68 10 content content NN 12483 68 11 served serve VBN 12483 68 12 by by IN 12483 68 13 digital digital JJ 12483 68 14 libraries library NNS 12483 68 15 can can MD 12483 68 16 be be VB 12483 68 17 accessed access VBN 12483 68 18 by by IN 12483 68 19 a a DT 12483 68 20 great great JJ 12483 68 21 number number NN 12483 68 22 of of IN 12483 68 23 users user NNS 12483 68 24 without without IN 12483 68 25 distance distance NN 12483 68 26 and and CC 12483 68 27 time time NN 12483 68 28 limits limit NNS 12483 68 29 . . . 12483 69 1 Therefore therefore RB 12483 69 2 , , , 12483 69 3 we -PRON- PRP 12483 69 4 propose propose VBP 12483 69 5 a a DT 12483 69 6 solution solution NN 12483 69 7 to to IN 12483 69 8 the the DT 12483 69 9 problem problem NN 12483 69 10 by by IN 12483 69 11 positioning position VBG 12483 69 12 digital digital JJ 12483 69 13 libraries library NNS 12483 69 14 as as IN 12483 69 15 automated automate VBN 12483 69 16 fact fact NN 12483 69 17 - - HYPH 12483 69 18 checking check VBG 12483 69 19 services service NNS 12483 69 20 , , , 12483 69 21 which which WDT 12483 69 22 label label VBP 12483 69 23 digital digital JJ 12483 69 24 news news NN 12483 69 25 content content NN 12483 69 26 as as IN 12483 69 27 fake fake JJ 12483 69 28 or or CC 12483 69 29 valid valid JJ 12483 69 30 as as RB 12483 69 31 soon soon RB 12483 69 32 as as IN 12483 69 33 or or CC 12483 69 34 before before IN 12483 69 35 it -PRON- PRP 12483 69 36 is be VBZ 12483 69 37 served serve VBN 12483 69 38 through through IN 12483 69 39 library library NN 12483 69 40 systems system NNS 12483 69 41 . . . 12483 70 1 The the DT 12483 70 2 main main JJ 12483 70 3 reason reason NN 12483 70 4 we -PRON- PRP 12483 70 5 associate associate VBP 12483 70 6 this this DT 12483 70 7 approach approach NN 12483 70 8 with with IN 12483 70 9 digital digital JJ 12483 70 10 libraries library NNS 12483 70 11 is be VBZ 12483 70 12 their -PRON- PRP$ 12483 70 13 access access NN 12483 70 14 to to IN 12483 70 15 a a DT 12483 70 16 wide wide JJ 12483 70 17 variety variety NN 12483 70 18 of of IN 12483 70 19 digital digital JJ 12483 70 20 content content NN 12483 70 21 which which WDT 12483 70 22 can can MD 12483 70 23 be be VB 12483 70 24 used use VBN 12483 70 25 to to TO 12483 70 26 train train VB 12483 70 27 the the DT 12483 70 28 proposed propose VBN 12483 70 29 mathematical mathematical JJ 12483 70 30 models model NNS 12483 70 31 , , , 12483 70 32 as as RB 12483 70 33 well well RB 12483 70 34 as as IN 12483 70 35 their -PRON- PRP$ 12483 70 36 role role NN 12483 70 37 in in IN 12483 70 38 the the DT 12483 70 39 society society NN 12483 70 40 as as IN 12483 70 41 the the DT 12483 70 42 publisher publisher NN 12483 70 43 of of IN 12483 70 44 trusted trusted JJ 12483 70 45 information information NN 12483 70 46 . . . 12483 71 1 To to IN 12483 71 2 this this DT 12483 71 3 end end NN 12483 71 4 , , , 12483 71 5 we -PRON- PRP 12483 71 6 develop develop VBP 12483 71 7 a a DT 12483 71 8 mathematical mathematical JJ 12483 71 9 model model NN 12483 71 10 that that WDT 12483 71 11 is be VBZ 12483 71 12 trained train VBN 12483 71 13 using use VBG 12483 71 14 existing exist VBG 12483 71 15 news news NN 12483 71 16 content content NN 12483 71 17 served serve VBN 12483 71 18 by by IN 12483 71 19 digital digital JJ 12483 71 20 libraries library NNS 12483 71 21 , , , 12483 71 22 and and CC 12483 71 23 capable capable JJ 12483 71 24 of of IN 12483 71 25 labelling label VBG 12483 71 26 news news NN 12483 71 27 content content NN 12483 71 28 as as IN 12483 71 29 fake fake JJ 12483 71 30 or or CC 12483 71 31 valid valid JJ 12483 71 32 with with IN 12483 71 33 unprecedented unprecedented JJ 12483 71 34 accuracy accuracy NN 12483 71 35 . . . 12483 72 1 The the DT 12483 72 2 proposed propose VBN 12483 72 3 solution solution NN 12483 72 4 uses use VBZ 12483 72 5 machine machine NN 12483 72 6 learning learning NN 12483 72 7 techniques technique NNS 12483 72 8 with with IN 12483 72 9 an an DT 12483 72 10 optimized optimized JJ 12483 72 11 set set NN 12483 72 12 of of IN 12483 72 13 extracted extract VBN 12483 72 14 features feature NNS 12483 72 15 and and CC 12483 72 16 annotated annotate VBD 12483 72 17 labels label NNS 12483 72 18 of of IN 12483 72 19 existing exist VBG 12483 72 20 digital digital JJ 12483 72 21 news news NN 12483 72 22 content content NN 12483 72 23 . . . 12483 73 1 Our -PRON- PRP$ 12483 73 2 study study NN 12483 73 3 mainly mainly RB 12483 73 4 contributes contribute VBZ 12483 73 5 ( ( -LRB- 12483 73 6 a a LS 12483 73 7 ) ) -RRB- 12483 73 8 a a DT 12483 73 9 new new JJ 12483 73 10 set set NN 12483 73 11 of of IN 12483 73 12 features feature NNS 12483 73 13 highly highly RB 12483 73 14 applicable applicable JJ 12483 73 15 for for IN 12483 73 16 agglutinative agglutinative JJ 12483 73 17 languages language NNS 12483 73 18 , , , 12483 73 19 ( ( -LRB- 12483 73 20 b b LS 12483 73 21 ) ) -RRB- 12483 73 22 the the DT 12483 73 23 first first JJ 12483 73 24 hybrid hybrid JJ 12483 73 25 model model NN 12483 73 26 combining combine VBG 12483 73 27 a a DT 12483 73 28 lexicon lexicon NN 12483 73 29 / / SYM 12483 73 30 dictionary- dictionary- NNP 12483 73 31 based base VBN 12483 73 32 approach approach NN 12483 73 33 with with IN 12483 73 34 machine machine NN 12483 73 35 learning learning NN 12483 73 36 methods method NNS 12483 73 37 to to TO 12483 73 38 detect detect VB 12483 73 39 fake fake JJ 12483 73 40 news news NN 12483 73 41 , , , 12483 73 42 and and CC 12483 73 43 ( ( -LRB- 12483 73 44 c c NN 12483 73 45 ) ) -RRB- 12483 73 46 a a DT 12483 73 47 benchmark benchmark JJ 12483 73 48 dataset dataset NN 12483 73 49 prepared prepare VBN 12483 73 50 in in IN 12483 73 51 Turkish Turkish NNP 12483 73 52 for for IN 12483 73 53 fake fake JJ 12483 73 54 news news NN 12483 73 55 detection detection NN 12483 73 56 . . . 12483 74 1 LITERATURE LITERATURE NNP 12483 74 2 REVIEW REVIEW NNP 12483 74 3 Contemporary Contemporary NNP 12483 74 4 studies study NNS 12483 74 5 have have VBP 12483 74 6 indicated indicate VBN 12483 74 7 that that IN 12483 74 8 social social JJ 12483 74 9 , , , 12483 74 10 economic economic JJ 12483 74 11 , , , 12483 74 12 and and CC 12483 74 13 political political JJ 12483 74 14 events event NNS 12483 74 15 in in IN 12483 74 16 recent recent JJ 12483 74 17 years year NNS 12483 74 18 , , , 12483 74 19 especially especially RB 12483 74 20 after after IN 12483 74 21 the the DT 12483 74 22 2016 2016 CD 12483 74 23 US US NNP 12483 74 24 presidential presidential JJ 12483 74 25 elections election NNS 12483 74 26 , , , 12483 74 27 are be VBP 12483 74 28 increasingly increasingly RB 12483 74 29 associated associate VBN 12483 74 30 with with IN 12483 74 31 the the DT 12483 74 32 concept concept NN 12483 74 33 of of IN 12483 74 34 fake fake JJ 12483 74 35 news.19 news.19 NNP 12483 74 36 Since since IN 12483 74 37 then then RB 12483 74 38 , , , 12483 74 39 fake fake JJ 12483 74 40 news news NN 12483 74 41 has have VBZ 12483 74 42 begun begin VBN 12483 74 43 to to TO 12483 74 44 be be VB 12483 74 45 used use VBN 12483 74 46 as as IN 12483 74 47 a a DT 12483 74 48 tool tool NN 12483 74 49 in in IN 12483 74 50 many many JJ 12483 74 51 domains domain NNS 12483 74 52 . . . 12483 75 1 On on IN 12483 75 2 the the DT 12483 75 3 other other JJ 12483 75 4 hand hand NN 12483 75 5 , , , 12483 75 6 researchers researcher NNS 12483 75 7 motivated motivate VBN 12483 75 8 by by IN 12483 75 9 finding find VBG 12483 75 10 automated automate VBN 12483 75 11 solutions solution NNS 12483 75 12 started start VBD 12483 75 13 to to TO 12483 75 14 make make VB 12483 75 15 use use NN 12483 75 16 of of IN 12483 75 17 machine machine NN 12483 75 18 learning learning NN 12483 75 19 , , , 12483 75 20 deep deep JJ 12483 75 21 learning learning NN 12483 75 22 , , , 12483 75 23 hybrid hybrid JJ 12483 75 24 models model NNS 12483 75 25 , , , 12483 75 26 and and CC 12483 75 27 other other JJ 12483 75 28 methodologies methodology NNS 12483 75 29 for for IN 12483 75 30 their -PRON- PRP$ 12483 75 31 solutions solution NNS 12483 75 32 . . . 12483 76 1 https://storyzy.com/ https://storyzy.com/ CD 12483 76 2 INFORMATION INFORMATION NNP 12483 76 3 TECHNOLOGY technology NN 12483 76 4 AND and CC 12483 76 5 LIBRARIES library NNS 12483 76 6 DECEMBER DECEMBER NNP 12483 76 7 2020 2020 CD 12483 76 8 AUTOMATED AUTOMATED NNP 12483 76 9 FAKE FAKE NNP 12483 76 10 NEWS NEWS NNP 12483 76 11 DETECTION detection VB 12483 76 12 IN in IN 12483 76 13 THE the DT 12483 76 14 AGE AGE NNP 12483 76 15 OF of IN 12483 76 16 DIGITAL DIGITAL NNP 12483 76 17 LIBRARIES library NNS 12483 76 18 | | NNP 12483 76 19 MERTOĞLU MERTOĞLU NNS 12483 76 20 AND and CC 12483 76 21 GENÇ genç NN 12483 76 22 5 5 CD 12483 76 23 Although although IN 12483 76 24 computational computational JJ 12483 76 25 deception deception NN 12483 76 26 detection detection NN 12483 76 27 studies study NNS 12483 76 28 applying apply VBG 12483 76 29 NLP NLP NNP 12483 76 30 ( ( -LRB- 12483 76 31 Natural Natural NNP 12483 76 32 Language Language NNP 12483 76 33 Processing Processing NNP 12483 76 34 ) ) -RRB- 12483 76 35 operations operation NNS 12483 76 36 are be VBP 12483 76 37 not not RB 12483 76 38 new new JJ 12483 76 39 , , , 12483 76 40 textual textual JJ 12483 76 41 deception deception NN 12483 76 42 in in IN 12483 76 43 the the DT 12483 76 44 context context NN 12483 76 45 of of IN 12483 76 46 text text NN 12483 76 47 - - HYPH 12483 76 48 based base VBN 12483 76 49 news news NN 12483 76 50 is be VBZ 12483 76 51 a a DT 12483 76 52 new new JJ 12483 76 53 topic topic NN 12483 76 54 for for IN 12483 76 55 the the DT 12483 76 56 field field NN 12483 76 57 of of IN 12483 76 58 journalism.20 journalism.20 NNP 12483 76 59 Accordingly accordingly RB 12483 76 60 , , , 12483 76 61 we -PRON- PRP 12483 76 62 believe believe VBP 12483 76 63 that that IN 12483 76 64 there there EX 12483 76 65 is be VBZ 12483 76 66 a a DT 12483 76 67 hidden hidden JJ 12483 76 68 body body NN 12483 76 69 language language NN 12483 76 70 of of IN 12483 76 71 news news NN 12483 76 72 text text NN 12483 76 73 , , , 12483 76 74 which which WDT 12483 76 75 has have VBZ 12483 76 76 linguistic linguistic JJ 12483 76 77 clues clue NNS 12483 76 78 indicating indicate VBG 12483 76 79 whether whether IN 12483 76 80 the the DT 12483 76 81 news news NN 12483 76 82 is be VBZ 12483 76 83 fake fake JJ 12483 76 84 or or CC 12483 76 85 not not RB 12483 76 86 . . . 12483 77 1 Thus thus RB 12483 77 2 , , , 12483 77 3 lexical lexical JJ 12483 77 4 , , , 12483 77 5 syntactic syntactic JJ 12483 77 6 , , , 12483 77 7 semantic semantic JJ 12483 77 8 , , , 12483 77 9 and and CC 12483 77 10 rhetorical rhetorical JJ 12483 77 11 analysis analysis NN 12483 77 12 when when WRB 12483 77 13 used use VBN 12483 77 14 with with IN 12483 77 15 machine machine NN 12483 77 16 learning learning NN 12483 77 17 and and CC 12483 77 18 deep deep JJ 12483 77 19 learning learning NN 12483 77 20 techniques technique NNS 12483 77 21 offers offer VBZ 12483 77 22 encouraging encourage VBG 12483 77 23 directions direction NNS 12483 77 24 . . . 12483 78 1 The the DT 12483 78 2 textual textual JJ 12483 78 3 deception deception NN 12483 78 4 spread spread VBD 12483 78 5 over over IN 12483 78 6 a a DT 12483 78 7 wide wide JJ 12483 78 8 spectrum spectrum NN 12483 78 9 and and CC 12483 78 10 the the DT 12483 78 11 studies study NNS 12483 78 12 have have VBP 12483 78 13 utilized utilize VBN 12483 78 14 many many JJ 12483 78 15 different different JJ 12483 78 16 techniques technique NNS 12483 78 17 . . . 12483 79 1 There there EX 12483 79 2 are be VBP 12483 79 3 some some DT 12483 79 4 prominent prominent JJ 12483 79 5 studies study NNS 12483 79 6 which which WDT 12483 79 7 took take VBD 12483 79 8 the the DT 12483 79 9 problem problem NN 12483 79 10 as as IN 12483 79 11 a a DT 12483 79 12 binary binary JJ 12483 79 13 classification classification NN 12483 79 14 problem problem NN 12483 79 15 utilizing utilize VBG 12483 79 16 linguistic linguistic JJ 12483 79 17 clues.21 clues.21 NNP 12483 79 18 Although although IN 12483 79 19 it -PRON- PRP 12483 79 20 is be VBZ 12483 79 21 still still RB 12483 79 22 early early JJ 12483 79 23 to to TO 12483 79 24 say say VB 12483 79 25 the the DT 12483 79 26 linguistic linguistic JJ 12483 79 27 characteristics characteristic NNS 12483 79 28 of of IN 12483 79 29 fake fake JJ 12483 79 30 news news NN 12483 79 31 are be VBP 12483 79 32 fully fully RB 12483 79 33 understood understand VBN 12483 79 34 , , , 12483 79 35 research research NN 12483 79 36 into into IN 12483 79 37 fake fake JJ 12483 79 38 - - HYPH 12483 79 39 news news NN 12483 79 40 detection detection NN 12483 79 41 in in IN 12483 79 42 English English NNP 12483 79 43 - - HYPH 12483 79 44 language language NN 12483 79 45 texts text NNS 12483 79 46 is be VBZ 12483 79 47 relatively relatively RB 12483 79 48 advanced advanced JJ 12483 79 49 compared compare VBN 12483 79 50 to to IN 12483 79 51 that that DT 12483 79 52 in in IN 12483 79 53 other other JJ 12483 79 54 languages language NNS 12483 79 55 . . . 12483 80 1 In in IN 12483 80 2 contrast contrast NN 12483 80 3 , , , 12483 80 4 agglutinative agglutinative JJ 12483 80 5 languages language NNS 12483 80 6 such such JJ 12483 80 7 as as IN 12483 80 8 Turkish turkish JJ 12483 80 9 have have VBP 12483 80 10 been be VBN 12483 80 11 little little RB 12483 80 12 researched research VBN 12483 80 13 when when WRB 12483 80 14 it -PRON- PRP 12483 80 15 comes come VBZ 12483 80 16 to to IN 12483 80 17 fake fake JJ 12483 80 18 news news NN 12483 80 19 detection detection NN 12483 80 20 . . . 12483 81 1 Agglutinative agglutinative JJ 12483 81 2 languages language NNS 12483 81 3 enable enable VBP 12483 81 4 the the DT 12483 81 5 construction construction NN 12483 81 6 of of IN 12483 81 7 words word NNS 12483 81 8 by by IN 12483 81 9 adding add VBG 12483 81 10 various various JJ 12483 81 11 morphemes morpheme NNS 12483 81 12 , , , 12483 81 13 which which WDT 12483 81 14 means mean VBZ 12483 81 15 that that IN 12483 81 16 words word NNS 12483 81 17 that that WDT 12483 81 18 are be VBP 12483 81 19 not not RB 12483 81 20 practically practically RB 12483 81 21 in in IN 12483 81 22 use use NN 12483 81 23 may may MD 12483 81 24 exist exist VB 12483 81 25 theoretically theoretically RB 12483 81 26 . . . 12483 82 1 For for IN 12483 82 2 example example NN 12483 82 3 , , , 12483 82 4 “ " `` 12483 82 5 gerek gerek JJ 12483 82 6 - - HYPH 12483 82 7 siz siz JJ 12483 82 8 - - HYPH 12483 82 9 leş leş NN 12483 82 10 - - HYPH 12483 82 11 tir tir NN 12483 82 12 - - HYPH 12483 82 13 ebil- ebil- NN 12483 82 14 ecek ecek NNP 12483 82 15 - - HYPH 12483 82 16 leri leri NN 12483 82 17 - - HYPH 12483 82 18 miz miz NN 12483 82 19 - - HYPH 12483 82 20 den den NNP 12483 82 21 - - HYPH 12483 82 22 dir dir NNP 12483 82 23 , , , 12483 82 24 ” " '' 12483 82 25 is be VBZ 12483 82 26 a a DT 12483 82 27 theoretically theoretically RB 12483 82 28 possible possible JJ 12483 82 29 word word NN 12483 82 30 that that DT 12483 82 31 means mean VBZ 12483 82 32 “ " `` 12483 82 33 it -PRON- PRP 12483 82 34 is be VBZ 12483 82 35 one one CD 12483 82 36 of of IN 12483 82 37 the the DT 12483 82 38 things thing NNS 12483 82 39 that that WDT 12483 82 40 we -PRON- PRP 12483 82 41 will will MD 12483 82 42 be be VB 12483 82 43 able able JJ 12483 82 44 to to TO 12483 82 45 make make VB 12483 82 46 redundant redundant JJ 12483 82 47 , , , 12483 82 48 ” " '' 12483 82 49 but but CC 12483 82 50 it -PRON- PRP 12483 82 51 is be VBZ 12483 82 52 not not RB 12483 82 53 a a DT 12483 82 54 practical practical JJ 12483 82 55 one one NN 12483 82 56 . . . 12483 83 1 Shu Shu NNP 12483 83 2 et et NNP 12483 83 3 al al NNP 12483 83 4 . . . 12483 84 1 classified classify VBD 12483 84 2 the the DT 12483 84 3 models model NNS 12483 84 4 for for IN 12483 84 5 the the DT 12483 84 6 detection detection NN 12483 84 7 of of IN 12483 84 8 fake fake JJ 12483 84 9 news news NN 12483 84 10 in in IN 12483 84 11 their -PRON- PRP$ 12483 84 12 study.22 study.22 NNP 12483 84 13 According accord VBG 12483 84 14 to to IN 12483 84 15 this this DT 12483 84 16 study study NN 12483 84 17 , , , 12483 84 18 the the DT 12483 84 19 automated automate VBN 12483 84 20 approaches approach NNS 12483 84 21 can can MD 12483 84 22 focus focus VB 12483 84 23 on on IN 12483 84 24 four four CD 12483 84 25 types type NNS 12483 84 26 of of IN 12483 84 27 attributes attribute NNS 12483 84 28 to to TO 12483 84 29 detect detect VB 12483 84 30 fake fake JJ 12483 84 31 news news NN 12483 84 32 : : : 12483 84 33 knowledge knowledge NN 12483 84 34 based base VBN 12483 84 35 , , , 12483 84 36 style style NN 12483 84 37 based base VBN 12483 84 38 , , , 12483 84 39 stance stance NN 12483 84 40 based base VBN 12483 84 41 , , , 12483 84 42 or or CC 12483 84 43 propagation propagation NN 12483 84 44 based base VBN 12483 84 45 . . . 12483 85 1 Among among IN 12483 85 2 these these DT 12483 85 3 , , , 12483 85 4 it -PRON- PRP 12483 85 5 can can MD 12483 85 6 be be VB 12483 85 7 said say VBN 12483 85 8 that that IN 12483 85 9 the the DT 12483 85 10 most most RBS 12483 85 11 useful useful JJ 12483 85 12 approaches approach NNS 12483 85 13 are be VBP 12483 85 14 the the DT 12483 85 15 ones one NNS 12483 85 16 which which WDT 12483 85 17 focus focus VBP 12483 85 18 on on IN 12483 85 19 the the DT 12483 85 20 textual textual JJ 12483 85 21 news news NN 12483 85 22 content content NN 12483 85 23 . . . 12483 86 1 Th Th NNP 12483 86 2 e e DT 12483 86 3 textual textual JJ 12483 86 4 content content NN 12483 86 5 can can MD 12483 86 6 be be VB 12483 86 7 studied study VBN 12483 86 8 by by IN 12483 86 9 an an DT 12483 86 10 automated automated JJ 12483 86 11 process process NN 12483 86 12 to to TO 12483 86 13 extract extract VB 12483 86 14 features feature NNS 12483 86 15 that that WDT 12483 86 16 can can MD 12483 86 17 be be VB 12483 86 18 very very RB 12483 86 19 helpful helpful JJ 12483 86 20 in in IN 12483 86 21 classifying classify VBG 12483 86 22 content content NN 12483 86 23 as as IN 12483 86 24 fake fake JJ 12483 86 25 or or CC 12483 86 26 valid valid JJ 12483 86 27 . . . 12483 87 1 Many many JJ 12483 87 2 scholars scholar NNS 12483 87 3 have have VBP 12483 87 4 tried try VBN 12483 87 5 to to TO 12483 87 6 build build VB 12483 87 7 models model NNS 12483 87 8 for for IN 12483 87 9 automatic automatic JJ 12483 87 10 detection detection NN 12483 87 11 and and CC 12483 87 12 prediction prediction NN 12483 87 13 of of IN 12483 87 14 fake fake JJ 12483 87 15 news news NN 12483 87 16 using use VBG 12483 87 17 machine machine NN 12483 87 18 learning learn VBG 12483 87 19 algorithms algorithm NNS 12483 87 20 , , , 12483 87 21 deep deep JJ 12483 87 22 learning learning NN 12483 87 23 algorithms algorithm NNS 12483 87 24 , , , 12483 87 25 and and CC 12483 87 26 other other JJ 12483 87 27 techniques technique NNS 12483 87 28 . . . 12483 88 1 These these DT 12483 88 2 scholars scholar NNS 12483 88 3 approach approach VBP 12483 88 4 the the DT 12483 88 5 detection detection NN 12483 88 6 of of IN 12483 88 7 fake fake JJ 12483 88 8 news news NN 12483 88 9 from from IN 12483 88 10 many many JJ 12483 88 11 different different JJ 12483 88 12 perspectives perspective NNS 12483 88 13 and and CC 12483 88 14 domains domain NNS 12483 88 15 . . . 12483 89 1 For for IN 12483 89 2 example example NN 12483 89 3 , , , 12483 89 4 in in IN 12483 89 5 one one CD 12483 89 6 of of IN 12483 89 7 the the DT 12483 89 8 studies study NNS 12483 89 9 , , , 12483 89 10 scientific scientific JJ 12483 89 11 news news NN 12483 89 12 and and CC 12483 89 13 conspiracy conspiracy NN 12483 89 14 news news NN 12483 89 15 were be VBD 12483 89 16 used.23 used.23 NNP 12483 89 17 In in IN 12483 89 18 Shu Shu NNP 12483 89 19 et et NNP 12483 89 20 al al NNP 12483 89 21 . . NNP 12483 89 22 ’s ’s NNP 12483 89 23 study study NN 12483 89 24 based base VBN 12483 89 25 on on IN 12483 89 26 credibility credibility NN 12483 89 27 of of IN 12483 89 28 news news NN 12483 89 29 , , , 12483 89 30 the the DT 12483 89 31 headlines headline NNS 12483 89 32 were be VBD 12483 89 33 used use VBN 12483 89 34 to to TO 12483 89 35 determine determine VB 12483 89 36 whether whether IN 12483 89 37 the the DT 12483 89 38 article article NN 12483 89 39 was be VBD 12483 89 40 clickbait clickbait JJ 12483 89 41 or or CC 12483 89 42 not not RB 12483 89 43 . . . 12483 90 1 In in IN 12483 90 2 another another DT 12483 90 3 study study NN 12483 90 4 , , , 12483 90 5 Reis Reis NNP 12483 90 6 et et NNP 12483 90 7 al al NNP 12483 90 8 . . . 12483 91 1 worked work VBN 12483 91 2 on on IN 12483 91 3 Buzzfeed Buzzfeed NNP 12483 91 4 articles article NNS 12483 91 5 linked link VBN 12483 91 6 to to IN 12483 91 7 the the DT 12483 91 8 2016 2016 CD 12483 91 9 US US NNP 12483 91 10 election election NN 12483 91 11 using use VBG 12483 91 12 machine machine NN 12483 91 13 learning learning NN 12483 91 14 techniques technique NNS 12483 91 15 with with IN 12483 91 16 a a DT 12483 91 17 supervised supervise VBN 12483 91 18 learning learn VBG 12483 91 19 approach.24 approach.24 NNP 12483 91 20 Studies Studies NNP 12483 91 21 which which WDT 12483 91 22 try try VBP 12483 91 23 to to TO 12483 91 24 detect detect VB 12483 91 25 satire satire NN 12483 91 26 and and CC 12483 91 27 sarcasm sarcasm NN 12483 91 28 can can MD 12483 91 29 be be VB 12483 91 30 attributed attribute VBN 12483 91 31 to to IN 12483 91 32 subcategories subcategorie NNS 12483 91 33 of of IN 12483 91 34 fake fake JJ 12483 91 35 news news NN 12483 91 36 detection.25 detection.25 NNP 12483 91 37 Our -PRON- PRP$ 12483 91 38 observation observation NN 12483 91 39 , , , 12483 91 40 in in IN 12483 91 41 line line NN 12483 91 42 with with IN 12483 91 43 the the DT 12483 91 44 general general JJ 12483 91 45 view view NN 12483 91 46 , , , 12483 91 47 is be VBZ 12483 91 48 that that DT 12483 91 49 satire satire NN 12483 91 50 is be VBZ 12483 91 51 not not RB 12483 91 52 always always RB 12483 91 53 recognizable recognizable JJ 12483 91 54 and and CC 12483 91 55 can can MD 12483 91 56 be be VB 12483 91 57 misunderstood misunderstand VBN 12483 91 58 for for IN 12483 91 59 real real JJ 12483 91 60 news.26 news.26 NNP 12483 91 61 For for IN 12483 91 62 this this DT 12483 91 63 reason reason NN 12483 91 64 , , , 12483 91 65 we -PRON- PRP 12483 91 66 included include VBD 12483 91 67 satirical satirical JJ 12483 91 68 news news NN 12483 91 69 in in IN 12483 91 70 our -PRON- PRP$ 12483 91 71 dataset dataset NN 12483 91 72 . . . 12483 92 1 It -PRON- PRP 12483 92 2 should should MD 12483 92 3 be be VB 12483 92 4 noted note VBN 12483 92 5 that that IN 12483 92 6 although although IN 12483 92 7 satire satire NN 12483 92 8 or or CC 12483 92 9 sarcasm sarcasm NN 12483 92 10 can can MD 12483 92 11 be be VB 12483 92 12 classified classify VBN 12483 92 13 by by IN 12483 92 14 automated automate VBN 12483 92 15 detection detection NN 12483 92 16 systems system NNS 12483 92 17 , , , 12483 92 18 experts expert NNS 12483 92 19 should should MD 12483 92 20 still still RB 12483 92 21 evaluate evaluate VB 12483 92 22 the the DT 12483 92 23 results result NNS 12483 92 24 of of IN 12483 92 25 the the DT 12483 92 26 classification classification NN 12483 92 27 . . . 12483 93 1 While while IN 12483 93 2 some some DT 12483 93 3 scholars scholar NNS 12483 93 4 used use VBD 12483 93 5 specific specific JJ 12483 93 6 models model NNS 12483 93 7 focusing focus VBG 12483 93 8 on on IN 12483 93 9 unique unique JJ 12483 93 10 characteristics characteristic NNS 12483 93 11 , , , 12483 93 12 some some DT 12483 93 13 others other NNS 12483 93 14 such such JJ 12483 93 15 as as IN 12483 93 16 Ruchansky Ruchansky NNP 12483 93 17 et et FW 12483 93 18 al al NNP 12483 93 19 . . . 12483 94 1 proposed propose VBN 12483 94 2 hybrid hybrid JJ 12483 94 3 deep deep JJ 12483 94 4 models model NNS 12483 94 5 for for IN 12483 94 6 fake fake JJ 12483 94 7 news news NN 12483 94 8 detection detection NN 12483 94 9 making make VBG 12483 94 10 use use NN 12483 94 11 of of IN 12483 94 12 multiple multiple JJ 12483 94 13 kinds kind NNS 12483 94 14 of of IN 12483 94 15 features feature NNS 12483 94 16 such such JJ 12483 94 17 as as IN 12483 94 18 temporal temporal JJ 12483 94 19 engagement engagement NN 12483 94 20 between between IN 12483 94 21 users user NNS 12483 94 22 and and CC 12483 94 23 news news NN 12483 94 24 articles article NNS 12483 94 25 over over IN 12483 94 26 time time NN 12483 94 27 and and CC 12483 94 28 generated generate VBD 12483 94 29 a a DT 12483 94 30 labelling labelling NN 12483 94 31 methodology methodology NN 12483 94 32 based base VBN 12483 94 33 on on IN 12483 94 34 those those DT 12483 94 35 features.27 features.27 NNP 12483 94 36 In in IN 12483 94 37 related related JJ 12483 94 38 studies study NNS 12483 94 39 , , , 12483 94 40 many many JJ 12483 94 41 features feature NNS 12483 94 42 such such JJ 12483 94 43 as as IN 12483 94 44 automatic automatic JJ 12483 94 45 extracted extract VBN 12483 94 46 features feature NNS 12483 94 47 , , , 12483 94 48 hand hand NN 12483 94 49 - - HYPH 12483 94 50 crafted craft VBN 12483 94 51 features feature NNS 12483 94 52 , , , 12483 94 53 social social JJ 12483 94 54 features feature NNS 12483 94 55 , , , 12483 94 56 network network NN 12483 94 57 information information NN 12483 94 58 , , , 12483 94 59 visual visual JJ 12483 94 60 features feature NNS 12483 94 61 , , , 12483 94 62 and and CC 12483 94 63 some some DT 12483 94 64 others other NNS 12483 94 65 such such JJ 12483 94 66 as as IN 12483 94 67 psycholinguistic psycholinguistic JJ 12483 94 68 features feature NNS 12483 94 69 , , , 12483 94 70 are be VBP 12483 94 71 applied apply VBN 12483 94 72 by by IN 12483 94 73 researchers.28 researchers.28 NNP 12483 94 74 In in IN 12483 94 75 this this DT 12483 94 76 work work NN 12483 94 77 , , , 12483 94 78 we -PRON- PRP 12483 94 79 focused focus VBD 12483 94 80 on on IN 12483 94 81 news news NN 12483 94 82 content content NN 12483 94 83 features feature NNS 12483 94 84 , , , 12483 94 85 however however RB 12483 94 86 the the DT 12483 94 87 social social JJ 12483 94 88 context context NN 12483 94 89 features feature NNS 12483 94 90 can can MD 12483 94 91 also also RB 12483 94 92 be be VB 12483 94 93 adapted adapt VBN 12483 94 94 using use VBG 12483 94 95 different different JJ 12483 94 96 tiers tier NNS 12483 94 97 such such JJ 12483 94 98 as as IN 12483 94 99 user user NN 12483 94 100 activity activity NN 12483 94 101 patterns pattern NNS 12483 94 102 , , , 12483 94 103 INFORMATION INFORMATION VBD 12483 94 104 TECHNOLOGY technology NN 12483 94 105 AND and CC 12483 94 106 LIBRARIES library NNS 12483 94 107 DECEMBER DECEMBER NNP 12483 94 108 2020 2020 CD 12483 94 109 AUTOMATED AUTOMATED NNP 12483 94 110 FAKE FAKE NNP 12483 94 111 NEWS NEWS NNP 12483 94 112 DETECTION detection VB 12483 94 113 IN in IN 12483 94 114 THE the DT 12483 94 115 AGE AGE NNP 12483 94 116 OF of IN 12483 94 117 DIGITAL DIGITAL NNP 12483 94 118 LIBRARIES library NNS 12483 94 119 | | NNP 12483 94 120 MERTOĞLU MERTOĞLU NNS 12483 94 121 AND and CC 12483 94 122 GENÇ genç VBP 12483 94 123 6 6 CD 12483 94 124 analysis analysis NN 12483 94 125 of of IN 12483 94 126 user user NN 12483 94 127 interaction interaction NN 12483 94 128 , , , 12483 94 129 profile profile NN 12483 94 130 metadata metadata NN 12483 94 131 , , , 12483 94 132 social social JJ 12483 94 133 network network NN 12483 94 134 / / SYM 12483 94 135 graph graph NN 12483 94 136 analysis analysis NN 12483 94 137 etc etc FW 12483 94 138 . . . 12483 95 1 to to TO 12483 95 2 extract extract VB 12483 95 3 features feature NNS 12483 95 4 . . . 12483 96 1 We -PRON- PRP 12483 96 2 also also RB 12483 96 3 have have VBP 12483 96 4 some some DT 12483 96 5 of of IN 12483 96 6 these these DT 12483 96 7 features feature NNS 12483 96 8 in in IN 12483 96 9 our -PRON- PRP$ 12483 96 10 data datum NNS 12483 96 11 but but CC 12483 96 12 not not RB 12483 96 13 having have VBG 12483 96 14 ground ground NN 12483 96 15 truth truth NN 12483 96 16 quantitatively quantitatively RB 12483 96 17 , , , 12483 96 18 we -PRON- PRP 12483 96 19 avoided avoid VBD 12483 96 20 using use VBG 12483 96 21 these these DT 12483 96 22 features feature NNS 12483 96 23 . . . 12483 97 1 METHODOLOGY METHODOLOGY NNP 12483 97 2 In in IN 12483 97 3 this this DT 12483 97 4 section section NN 12483 97 5 , , , 12483 97 6 we -PRON- PRP 12483 97 7 present present VBP 12483 97 8 our -PRON- PRP$ 12483 97 9 motivation motivation NN 12483 97 10 for for IN 12483 97 11 this this DT 12483 97 12 work work NN 12483 97 13 which which WDT 12483 97 14 we -PRON- PRP 12483 97 15 visualized visualize VBD 12483 97 16 in in IN 12483 97 17 a a DT 12483 97 18 framework framework NN 12483 97 19 and and CC 12483 97 20 named name VBN 12483 97 21 Global Global NNP 12483 97 22 Library Library NNP 12483 97 23 and and CC 12483 97 24 Information Information NNP 12483 97 25 Science Science NNP 12483 97 26 ( ( -LRB- 12483 97 27 GLIS_1.0 glis_1.0 CD 12483 97 28 ) ) -RRB- 12483 97 29 . . . 12483 98 1 Subsequently subsequently RB 12483 98 2 , , , 12483 98 3 we -PRON- PRP 12483 98 4 discuss discuss VBP 12483 98 5 the the DT 12483 98 6 construction construction NN 12483 98 7 of of IN 12483 98 8 the the DT 12483 98 9 automated automate VBN 12483 98 10 detection detection NN 12483 98 11 system system NN 12483 98 12 as as IN 12483 98 13 the the DT 12483 98 14 key key JJ 12483 98 15 element element NN 12483 98 16 of of IN 12483 98 17 the the DT 12483 98 18 GLIS_1.0 glis_1.0 CD 12483 98 19 framework framework NN 12483 98 20 . . . 12483 99 1 We -PRON- PRP 12483 99 2 explain explain VBP 12483 99 3 the the DT 12483 99 4 framework framework NN 12483 99 5 , , , 12483 99 6 model model NN 12483 99 7 , , , 12483 99 8 dataset dataset NN 12483 99 9 , , , 12483 99 10 features feature NNS 12483 99 11 , , , 12483 99 12 and and CC 12483 99 13 the the DT 12483 99 14 techniques technique NNS 12483 99 15 used use VBN 12483 99 16 in in IN 12483 99 17 this this DT 12483 99 18 section section NN 12483 99 19 . . . 12483 100 1 Framework framework VB 12483 100 2 The the DT 12483 100 3 main main JJ 12483 100 4 structure structure NN 12483 100 5 of of IN 12483 100 6 the the DT 12483 100 7 proposed propose VBN 12483 100 8 framework framework NN 12483 100 9 is be VBZ 12483 100 10 shown show VBN 12483 100 11 in in IN 12483 100 12 figure figure NN 12483 100 13 2 2 CD 12483 100 14 . . . 12483 101 1 This this DT 12483 101 2 framework framework NN 12483 101 3 consists consist VBZ 12483 101 4 of of IN 12483 101 5 highly highly RB 12483 101 6 cohesive cohesive JJ 12483 101 7 but but CC 12483 101 8 flexible flexible JJ 12483 101 9 layers layer NNS 12483 101 10 . . . 12483 102 1 Figure figure NN 12483 102 2 2 2 CD 12483 102 3 . . . 12483 103 1 The the DT 12483 103 2 GLIS_1.0 glis_1.0 CD 12483 103 3 framework framework NN 12483 103 4 main main JJ 12483 103 5 structure structure NN 12483 103 6 . . . 12483 104 1 In in IN 12483 104 2 the the DT 12483 104 3 presentation presentation NN 12483 104 4 layer layer NN 12483 104 5 one one PRP 12483 104 6 can can MD 12483 104 7 find find VB 12483 104 8 the the DT 12483 104 9 different different JJ 12483 104 10 sources source NNS 12483 104 11 of of IN 12483 104 12 news news NN 12483 104 13 that that WDT 12483 104 14 are be VBP 12483 104 15 publicly publicly RB 12483 104 16 available available JJ 12483 104 17 . . . 12483 105 1 These these DT 12483 105 2 sources source NNS 12483 105 3 can can MD 12483 105 4 be be VB 12483 105 5 accessed access VBN 12483 105 6 directly directly RB 12483 105 7 using use VBG 12483 105 8 their -PRON- PRP$ 12483 105 9 websites website NNS 12483 105 10 or or CC 12483 105 11 can can MD 12483 105 12 be be VB 12483 105 13 searched search VBN 12483 105 14 for for IN 12483 105 15 via via IN 12483 105 16 search search NN 12483 105 17 engines engine NNS 12483 105 18 . . . 12483 106 1 The the DT 12483 106 2 news news NN 12483 106 3 is be VBZ 12483 106 4 received receive VBN 12483 106 5 by by IN 12483 106 6 fact fact NN 12483 106 7 - - HYPH 12483 106 8 checking check VBG 12483 106 9 organizations organization NNS 12483 106 10 which which WDT 12483 106 11 classify classify VBP 12483 106 12 them -PRON- PRP 12483 106 13 manually manually RB 12483 106 14 , , , 12483 106 15 digital digital JJ 12483 106 16 libraries library NNS 12483 106 17 which which WDT 12483 106 18 archives archive VBZ 12483 106 19 and and CC 12483 106 20 serves serve VBZ 12483 106 21 them -PRON- PRP 12483 106 22 , , , 12483 106 23 and and CC 12483 106 24 automated automate VBD 12483 106 25 detection detection NN 12483 106 26 systems system NNS 12483 106 27 ( ( -LRB- 12483 106 28 ADS ADS NNP 12483 106 29 ) ) -RRB- 12483 106 30 which which WDT 12483 106 31 classify classify VBP 12483 106 32 them -PRON- PRP 12483 106 33 automatically automatically RB 12483 106 34 . . . 12483 107 1 Digital digital JJ 12483 107 2 libraries library NNS 12483 107 3 work work VBP 12483 107 4 together together RB 12483 107 5 with with IN 12483 107 6 fact fact NN 12483 107 7 - - HYPH 12483 107 8 checking check VBG 12483 107 9 organizations organization NNS 12483 107 10 and and CC 12483 107 11 ADSs ads NNS 12483 107 12 to to TO 12483 107 13 present present VB 12483 107 14 clean clean JJ 12483 107 15 and and CC 12483 107 16 valid valid JJ 12483 107 17 news news NN 12483 107 18 to to IN 12483 107 19 the the DT 12483 107 20 public public NN 12483 107 21 . . . 12483 108 1 Moreover moreover RB 12483 108 2 , , , 12483 108 3 search search NN 12483 108 4 engines engine NNS 12483 108 5 use use VBP 12483 108 6 digital digital JJ 12483 108 7 libraries library NNS 12483 108 8 systems system NNS 12483 108 9 to to TO 12483 108 10 label label VB 12483 108 11 their -PRON- PRP$ 12483 108 12 results result NNS 12483 108 13 as as IN 12483 108 14 fake fake JJ 12483 108 15 or or CC 12483 108 16 valid valid JJ 12483 108 17 . . . 12483 109 1 INFORMATION INFORMATION NNP 12483 109 2 TECHNOLOGY technology NN 12483 109 3 AND and CC 12483 109 4 LIBRARIES library NNS 12483 109 5 DECEMBER DECEMBER NNP 12483 109 6 2020 2020 CD 12483 109 7 AUTOMATED AUTOMATED NNP 12483 109 8 FAKE FAKE NNP 12483 109 9 NEWS NEWS NNP 12483 109 10 DETECTION detection VB 12483 109 11 IN in IN 12483 109 12 THE the DT 12483 109 13 AGE AGE NNP 12483 109 14 OF of IN 12483 109 15 DIGITAL DIGITAL NNP 12483 109 16 LIBRARIES library NNS 12483 109 17 | | NNP 12483 109 18 MERTOĞLU MERTOĞLU NNS 12483 109 19 AND and CC 12483 109 20 GENÇ genç VBP 12483 109 21 7 7 CD 12483 109 22 Fact fact NN 12483 109 23 - - HYPH 12483 109 24 checking check VBG 12483 109 25 organizations organization NNS 12483 109 26 should should MD 12483 109 27 also also RB 12483 109 28 benefit benefit VB 12483 109 29 from from IN 12483 109 30 the the DT 12483 109 31 output output NN 12483 109 32 of of IN 12483 109 33 ADSs ads NNS 12483 109 34 , , , 12483 109 35 as as RB 12483 109 36 instead instead RB 12483 109 37 of of IN 12483 109 38 manually manually RB 12483 109 39 checking check VBG 12483 109 40 heaps heap NNS 12483 109 41 of of IN 12483 109 42 news news NN 12483 109 43 content content NN 12483 109 44 , , , 12483 109 45 they -PRON- PRP 12483 109 46 could could MD 12483 109 47 now now RB 12483 109 48 focus focus VB 12483 109 49 on on IN 12483 109 50 news news NN 12483 109 51 labeled label VBN 12483 109 52 as as IN 12483 109 53 potentially potentially RB 12483 109 54 fake fake JJ 12483 109 55 by by IN 12483 109 56 an an DT 12483 109 57 ADS ADS NNP 12483 109 58 . . . 12483 110 1 Through through IN 12483 110 2 GLIS GLIS NNP 12483 110 3 , , , 12483 110 4 ADSs ads NNS 12483 110 5 make make VBP 12483 110 6 the the DT 12483 110 7 life life NN 12483 110 8 of of IN 12483 110 9 fact fact NN 12483 110 10 - - HYPH 12483 110 11 checking check VBG 12483 110 12 organizations organization NNS 12483 110 13 and and CC 12483 110 14 digital digital JJ 12483 110 15 libraries library NNS 12483 110 16 much much RB 12483 110 17 easier easy JJR 12483 110 18 , , , 12483 110 19 all all PDT 12483 110 20 the the DT 12483 110 21 while while IN 12483 110 22 increasing increase VBG 12483 110 23 the the DT 12483 110 24 quality quality NN 12483 110 25 of of IN 12483 110 26 news news NN 12483 110 27 served serve VBD 12483 110 28 to to IN 12483 110 29 the the DT 12483 110 30 public public NN 12483 110 31 . . . 12483 111 1 Considering consider VBG 12483 111 2 this this DT 12483 111 3 is be VBZ 12483 111 4 a a DT 12483 111 5 high high JJ 12483 111 6 - - HYPH 12483 111 7 level level NN 12483 111 8 overview overview NN 12483 111 9 of of IN 12483 111 10 a a DT 12483 111 11 structure structure NN 12483 111 12 given give VBN 12483 111 13 in in IN 12483 111 14 figure figure NN 12483 111 15 2 2 CD 12483 111 16 , , , 12483 111 17 there there EX 12483 111 18 may may MD 12483 111 19 be be VB 12483 111 20 many many JJ 12483 111 21 other other JJ 12483 111 22 components component NNS 12483 111 23 , , , 12483 111 24 mechanisms mechanism NNS 12483 111 25 , , , 12483 111 26 or or CC 12483 111 27 layers layer NNS 12483 111 28 , , , 12483 111 29 but but CC 12483 111 30 the the DT 12483 111 31 key key JJ 12483 111 32 elements element NNS 12483 111 33 of of IN 12483 111 34 this this DT 12483 111 35 structure structure NN 12483 111 36 are be VBP 12483 111 37 automated automate VBN 12483 111 38 detection detection NN 12483 111 39 systems system NNS 12483 111 40 and and CC 12483 111 41 the the DT 12483 111 42 digital digital JJ 12483 111 43 libraries library NNS 12483 111 44 . . . 12483 112 1 A a DT 12483 112 2 critical critical JJ 12483 112 3 approach approach NN 12483 112 4 to to IN 12483 112 5 this this DT 12483 112 6 framework framework NN 12483 112 7 can can MD 12483 112 8 be be VB 12483 112 9 why why WRB 12483 112 10 we -PRON- PRP 12483 112 11 need need VBP 12483 112 12 such such PDT 12483 112 13 an an DT 12483 112 14 authority authority NN 12483 112 15 mechanism mechanism NN 12483 112 16 . . . 12483 113 1 The the DT 12483 113 2 answer answer NN 12483 113 3 will will MD 12483 113 4 be be VB 12483 113 5 quite quite RB 12483 113 6 simple simple JJ 12483 113 7 , , , 12483 113 8 technological technological JJ 12483 113 9 progress progress NN 12483 113 10 is be VBZ 12483 113 11 not not RB 12483 113 12 the the DT 12483 113 13 only only JJ 12483 113 14 solution solution NN 12483 113 15 . . . 12483 114 1 On on IN 12483 114 2 the the DT 12483 114 3 contrary contrary NN 12483 114 4 , , , 12483 114 5 tech tech NN 12483 114 6 giants giant NNS 12483 114 7 have have VBP 12483 114 8 already already RB 12483 114 9 been be VBN 12483 114 10 subject subject JJ 12483 114 11 to to IN 12483 114 12 regulatory regulatory JJ 12483 114 13 scrutiny scrutiny NN 12483 114 14 for for IN 12483 114 15 how how WRB 12483 114 16 they -PRON- PRP 12483 114 17 handle handle VBP 12483 114 18 personal personal JJ 12483 114 19 information.29 information.29 NNP 12483 114 20 Also also RB 12483 114 21 , , , 12483 114 22 their -PRON- PRP$ 12483 114 23 policy policy NN 12483 114 24 related relate VBN 12483 114 25 to to IN 12483 114 26 political political JJ 12483 114 27 ads ad NNS 12483 114 28 has have VBZ 12483 114 29 been be VBN 12483 114 30 questioned question VBN 12483 114 31 . . . 12483 115 1 Furthermore furthermore RB 12483 115 2 , , , 12483 115 3 they -PRON- PRP 12483 115 4 are be VBP 12483 115 5 often often RB 12483 115 6 blamed blame VBN 12483 115 7 for for IN 12483 115 8 failing fail VBG 12483 115 9 to to TO 12483 115 10 fight fight VB 12483 115 11 fake fake JJ 12483 115 12 news news NN 12483 115 13 . . . 12483 116 1 Indeed indeed RB 12483 116 2 , , , 12483 116 3 there there EX 12483 116 4 is be VBZ 12483 116 5 an an DT 12483 116 6 urgent urgent JJ 12483 116 7 need need NN 12483 116 8 for for IN 12483 116 9 a a DT 12483 116 10 global global JJ 12483 116 11 action action NN 12483 116 12 more more RBR 12483 116 13 than than IN 12483 116 14 ever ever RB 12483 116 15 . . . 12483 117 1 Digital digital JJ 12483 117 2 libraries library NNS 12483 117 3 are be VBP 12483 117 4 much much RB 12483 117 5 more more JJR 12483 117 6 than than IN 12483 117 7 a a DT 12483 117 8 technological technological JJ 12483 117 9 advancement advancement NN 12483 117 10 . . . 12483 118 1 Hence hence RB 12483 118 2 , , , 12483 118 3 they -PRON- PRP 12483 118 4 should should MD 12483 118 5 be be VB 12483 118 6 considered consider VBN 12483 118 7 as as IN 12483 118 8 institutions institution NNS 12483 118 9 or or CC 12483 118 10 services service NNS 12483 118 11 which which WDT 12483 118 12 can can MD 12483 118 13 be be VB 12483 118 14 a a DT 12483 118 15 great great JJ 12483 118 16 authority authority NN 12483 118 17 service service NN 12483 118 18 to to TO 12483 118 19 provide provide VB 12483 118 20 news news NN 12483 118 21 to to IN 12483 118 22 society society NN 12483 118 23 since since IN 12483 118 24 the the DT 12483 118 25 printed print VBN 12483 118 26 media media NN 12483 118 27 disappears disappear VBZ 12483 118 28 day day NN 12483 118 29 by by IN 12483 118 30 day day NN 12483 118 31 . . . 12483 119 1 The the DT 12483 119 2 threats threat NNS 12483 119 3 caused cause VBN 12483 119 4 by by IN 12483 119 5 fake fake JJ 12483 119 6 news news NN 12483 119 7 are be VBP 12483 119 8 real real JJ 12483 119 9 and and CC 12483 119 10 dangerous dangerous JJ 12483 119 11 , , , 12483 119 12 but but CC 12483 119 13 only only RB 12483 119 14 recently recently RB 12483 119 15 have have VBP 12483 119 16 researchers researcher NNS 12483 119 17 from from IN 12483 119 18 different different JJ 12483 119 19 disciplines discipline NNS 12483 119 20 been be VBN 12483 119 21 trying try VBG 12483 119 22 to to TO 12483 119 23 find find VB 12483 119 24 possible possible JJ 12483 119 25 solutions solution NNS 12483 119 26 such such JJ 12483 119 27 as as IN 12483 119 28 educational educational JJ 12483 119 29 , , , 12483 119 30 technological technological JJ 12483 119 31 , , , 12483 119 32 regulatory regulatory JJ 12483 119 33 , , , 12483 119 34 or or CC 12483 119 35 political political JJ 12483 119 36 . . . 12483 120 1 Digital digital JJ 12483 120 2 librarianship librarianship NN 12483 120 3 can can MD 12483 120 4 be be VB 12483 120 5 the the DT 12483 120 6 intersection intersection NN 12483 120 7 of of IN 12483 120 8 all all PDT 12483 120 9 these these DT 12483 120 10 solutions solution NNS 12483 120 11 for for IN 12483 120 12 promoting promote VBG 12483 120 13 information information NN 12483 120 14 / / SYM 12483 120 15 media medium NNS 12483 120 16 literacy literacy NN 12483 120 17 . . . 12483 121 1 Hence hence RB 12483 121 2 , , , 12483 121 3 digital digital JJ 12483 121 4 librarianship librarianship NN 12483 121 5 will will MD 12483 121 6 make make VB 12483 121 7 use use NN 12483 121 8 of of IN 12483 121 9 many many JJ 12483 121 10 automated automate VBN 12483 121 11 detection detection NN 12483 121 12 systems system NNS 12483 121 13 ( ( -LRB- 12483 121 14 ADS ADS NNP 12483 121 15 ) ) -RRB- 12483 121 16 to to TO 12483 121 17 serve serve VB 12483 121 18 qualified qualified JJ 12483 121 19 news news NN 12483 121 20 . . . 12483 122 1 In in IN 12483 122 2 the the DT 12483 122 3 following following JJ 12483 122 4 section section NN 12483 122 5 , , , 12483 122 6 we -PRON- PRP 12483 122 7 discuss discuss VBP 12483 122 8 ADS ADS NNP 12483 122 9 in in IN 12483 122 10 detail detail NN 12483 122 11 . . . 12483 123 1 Model Model NNP 12483 123 2 An an DT 12483 123 3 overview overview NN 12483 123 4 of of IN 12483 123 5 our -PRON- PRP$ 12483 123 6 model model NN 12483 123 7 of of IN 12483 123 8 automated automate VBN 12483 123 9 detection detection NN 12483 123 10 system system NN 12483 123 11 solution solution NN 12483 123 12 which which WDT 12483 123 13 is be VBZ 12483 123 14 very very RB 12483 123 15 critical critical JJ 12483 123 16 for for IN 12483 123 17 the the DT 12483 123 18 framework framework NN 12483 123 19 is be VBZ 12483 123 20 shown show VBN 12483 123 21 in in IN 12483 123 22 figure figure NN 12483 123 23 3 3 CD 12483 123 24 . . . 12483 124 1 Our -PRON- PRP$ 12483 124 2 fake fake JJ 12483 124 3 news news NN 12483 124 4 detection detection NN 12483 124 5 model model NN 12483 124 6 consists consist VBZ 12483 124 7 of of IN 12483 124 8 two two CD 12483 124 9 phases phase NNS 12483 124 10 . . . 12483 125 1 First first RB 12483 125 2 is be VBZ 12483 125 3 the the DT 12483 125 4 Language Language NNP 12483 125 5 Model Model NNP 12483 125 6 / / SYM 12483 125 7 Lexicon Lexicon NNP 12483 125 8 Generation Generation NNP 12483 125 9 and and CC 12483 125 10 the the DT 12483 125 11 second second JJ 12483 125 12 is be VBZ 12483 125 13 Machine Machine NNP 12483 125 14 Learning Learning NNP 12483 125 15 Integration Integration NNP 12483 125 16 . . . 12483 126 1 In in IN 12483 126 2 this this DT 12483 126 3 work work NN 12483 126 4 , , , 12483 126 5 we -PRON- PRP 12483 126 6 used use VBD 12483 126 7 machine machine NN 12483 126 8 learning learn VBG 12483 126 9 algorithms algorithm NNS 12483 126 10 via via IN 12483 126 11 supervised supervise VBN 12483 126 12 learning learning NN 12483 126 13 techniques technique NNS 12483 126 14 which which WDT 12483 126 15 learn learn VBP 12483 126 16 from from IN 12483 126 17 labeled label VBN 12483 126 18 news news NN 12483 126 19 data datum NNS 12483 126 20 ( ( -LRB- 12483 126 21 training training NN 12483 126 22 ) ) -RRB- 12483 126 23 and and CC 12483 126 24 helps help VBZ 12483 126 25 us -PRON- PRP 12483 126 26 to to TO 12483 126 27 predict predict VB 12483 126 28 outcomes outcome NNS 12483 126 29 for for IN 12483 126 30 unforeseen unforeseen JJ 12483 126 31 news news NNP 12483 126 32 data datum NNS 12483 126 33 ( ( -LRB- 12483 126 34 test test NN 12483 126 35 ) ) -RRB- 12483 126 36 . . . 12483 127 1 Dataset dataset FW 12483 127 2 We -PRON- PRP 12483 127 3 collected collect VBD 12483 127 4 our -PRON- PRP$ 12483 127 5 data datum NNS 12483 127 6 from from IN 12483 127 7 three three CD 12483 127 8 sources source NNS 12483 127 9 : : : 12483 127 10 • • VB 12483 127 11 The the DT 12483 127 12 primary primary JJ 12483 127 13 source source NN 12483 127 14 is be VBZ 12483 127 15 the the DT 12483 127 16 GDELT GDELT NNP 12483 127 17 ( ( -LRB- 12483 127 18 Global Global NNP 12483 127 19 Database Database NNP 12483 127 20 of of IN 12483 127 21 Events Events NNPS 12483 127 22 , , , 12483 127 23 Language Language NNP 12483 127 24 and and CC 12483 127 25 Tone Tone NNP 12483 127 26 ) ) -RRB- 12483 127 27 Project Project NNP 12483 127 28 ( ( -LRB- 12483 127 29 https://www.gdeltproject.org/ https://www.gdeltproject.org/ NNP 12483 127 30 ) ) -RRB- 12483 127 31 , , , 12483 127 32 a a DT 12483 127 33 massive massive JJ 12483 127 34 global global JJ 12483 127 35 news news NN 12483 127 36 media medium NNS 12483 127 37 archive archive JJ 12483 127 38 offering offer VBG 12483 127 39 free free JJ 12483 127 40 access access NN 12483 127 41 to to IN 12483 127 42 news news NN 12483 127 43 text text NN 12483 127 44 metadata metadata NN 12483 127 45 for for IN 12483 127 46 researchers researcher NNS 12483 127 47 worldwide worldwide RB 12483 127 48 . . . 12483 128 1 It -PRON- PRP 12483 128 2 can can MD 12483 128 3 almost almost RB 12483 128 4 be be VB 12483 128 5 considered consider VBN 12483 128 6 a a DT 12483 128 7 digital digital JJ 12483 128 8 library library NN 12483 128 9 of of IN 12483 128 10 news news NN 12483 128 11 in in IN 12483 128 12 its -PRON- PRP$ 12483 128 13 own own JJ 12483 128 14 right right NN 12483 128 15 . . . 12483 129 1 However however RB 12483 129 2 , , , 12483 129 3 GDELT GDELT NNP 12483 129 4 does do VBZ 12483 129 5 not not RB 12483 129 6 provide provide VB 12483 129 7 the the DT 12483 129 8 actual actual JJ 12483 129 9 news news NN 12483 129 10 text text NN 12483 129 11 and and CC 12483 129 12 only only RB 12483 129 13 serves serve VBZ 12483 129 14 processed process VBN 12483 129 15 metadata metadata NN 12483 129 16 along along IN 12483 129 17 with with IN 12483 129 18 the the DT 12483 129 19 URL url NN 12483 129 20 of of IN 12483 129 21 the the DT 12483 129 22 news news NN 12483 129 23 item item NN 12483 129 24 . . . 12483 130 1 GDELT GDELT NNP 12483 130 2 normally normally RB 12483 130 3 does do VBZ 12483 130 4 not not RB 12483 130 5 check check VB 12483 130 6 for for IN 12483 130 7 the the DT 12483 130 8 validity validity NN 12483 130 9 of of IN 12483 130 10 any any DT 12483 130 11 news news NN 12483 130 12 items item NNS 12483 130 13 . . . 12483 131 1 However however RB 12483 131 2 , , , 12483 131 3 we -PRON- PRP 12483 131 4 have have VBP 12483 131 5 only only RB 12483 131 6 used use VBN 12483 131 7 news news NN 12483 131 8 from from IN 12483 131 9 approved approve VBN 12483 131 10 news news NN 12483 131 11 agencies agency NNS 12483 131 12 and and CC 12483 131 13 completely completely RB 12483 131 14 ignored ignore VBN 12483 131 15 news news NN 12483 131 16 from from IN 12483 131 17 local local JJ 12483 131 18 and and CC 12483 131 19 lesser less JJR 12483 131 20 - - HYPH 12483 131 21 known know VBN 12483 131 22 sources source NNS 12483 131 23 to to TO 12483 131 24 maximize maximize VB 12483 131 25 the the DT 12483 131 26 validity validity NN 12483 131 27 of of IN 12483 131 28 the the DT 12483 131 29 news news NN 12483 131 30 we -PRON- PRP 12483 131 31 have have VBP 12483 131 32 automatically automatically RB 12483 131 33 obtained obtain VBN 12483 131 34 through through IN 12483 131 35 GDELT GDELT NNP 12483 131 36 . . . 12483 132 1 Moreover moreover RB 12483 132 2 , , , 12483 132 3 we -PRON- PRP 12483 132 4 have have VBP 12483 132 5 post post VBN 12483 132 6 - - VBN 12483 132 7 processed process VBN 12483 132 8 the the DT 12483 132 9 obtained obtain VBN 12483 132 10 texts text NNS 12483 132 11 by by IN 12483 132 12 cross cross JJ 12483 132 13 - - JJ 12483 132 14 validating validate VBG 12483 132 15 with with IN 12483 132 16 teyit.org teyit.org CD 12483 132 17 data datum NNS 12483 132 18 to to TO 12483 132 19 clean clean VB 12483 132 20 any any DT 12483 132 21 potential potential JJ 12483 132 22 fake fake JJ 12483 132 23 news news NN 12483 132 24 obtained obtain VBN 12483 132 25 through through IN 12483 132 26 GDELT GDELT NNP 12483 132 27 links link NNS 12483 132 28 . . . 12483 133 1 https://www.gdeltproject.org/ https://www.gdeltproject.org/ NNP 12483 133 2 INFORMATION INFORMATION NNP 12483 133 3 TECHNOLOGY TECHNOLOGY NNP 12483 133 4 AND and CC 12483 133 5 LIBRARIES library NNS 12483 133 6 DECEMBER DECEMBER NNP 12483 133 7 2020 2020 CD 12483 133 8 AUTOMATED AUTOMATED NNP 12483 133 9 FAKE FAKE NNP 12483 133 10 NEWS NEWS NNP 12483 133 11 DETECTION detection VB 12483 133 12 IN in IN 12483 133 13 THE the DT 12483 133 14 AGE AGE NNP 12483 133 15 OF of IN 12483 133 16 DIGITAL DIGITAL NNP 12483 133 17 LIBRARIES library NNS 12483 133 18 | | NNP 12483 133 19 MERTOĞLU MERTOĞLU NNS 12483 133 20 AND and CC 12483 133 21 GENÇ genç NN 12483 133 22 8 8 CD 12483 133 23 Figure figure NN 12483 133 24 3 3 CD 12483 133 25 . . . 12483 134 1 Integrated integrate VBN 12483 134 2 fake fake JJ 12483 134 3 news news NN 12483 134 4 detection detection NN 12483 134 5 model model NN 12483 134 6 with with IN 12483 134 7 main main JJ 12483 134 8 phases phase NNS 12483 134 9 combining combine VBG 12483 134 10 language language NN 12483 134 11 - - HYPH 12483 134 12 model model NN 12483 134 13 based base VBN 12483 134 14 approach approach NN 12483 134 15 with with IN 12483 134 16 machine machine NN 12483 134 17 learning learning NN 12483 134 18 approach approach NN 12483 134 19 . . . 12483 135 1 • • VB 12483 135 2 The the DT 12483 135 3 second second JJ 12483 135 4 source source NN 12483 135 5 is be VBZ 12483 135 6 teyit.org teyit.org ADD 12483 135 7 which which WDT 12483 135 8 is be VBZ 12483 135 9 a a DT 12483 135 10 fact fact NN 12483 135 11 - - HYPH 12483 135 12 checking check VBG 12483 135 13 organization organization NN 12483 135 14 based base VBN 12483 135 15 in in IN 12483 135 16 Turkey Turkey NNP 12483 135 17 , , , 12483 135 18 compliant compliant JJ 12483 135 19 to to IN 12483 135 20 the the DT 12483 135 21 principles principle NNS 12483 135 22 of of IN 12483 135 23 IFCN IFCN NNP 12483 135 24 ( ( -LRB- 12483 135 25 International International NNP 12483 135 26 Fact Fact NNP 12483 135 27 - - HYPH 12483 135 28 Checking Checking NNP 12483 135 29 Network Network NNP 12483 135 30 ) ) -RRB- 12483 135 31 aiming aim VBG 12483 135 32 to to TO 12483 135 33 prevent prevent VB 12483 135 34 spreading spreading NN 12483 135 35 of of IN 12483 135 36 false false JJ 12483 135 37 information information NN 12483 135 38 through through IN 12483 135 39 online online JJ 12483 135 40 channels channel NNS 12483 135 41 . . . 12483 136 1 Manually manually RB 12483 136 2 analyzing analyze VBG 12483 136 3 each each DT 12483 136 4 INFORMATION INFORMATION NNP 12483 136 5 TECHNOLOGY technology NN 12483 136 6 AND and CC 12483 136 7 LIBRARIES library NNS 12483 136 8 DECEMBER DECEMBER NNP 12483 136 9 2020 2020 CD 12483 136 10 AUTOMATED AUTOMATED NNP 12483 136 11 FAKE FAKE NNP 12483 136 12 NEWS NEWS NNP 12483 136 13 DETECTION detection VB 12483 136 14 IN in IN 12483 136 15 THE the DT 12483 136 16 AGE AGE NNP 12483 136 17 OF of IN 12483 136 18 DIGITAL DIGITAL NNP 12483 136 19 LIBRARIES library NNS 12483 136 20 | | NNP 12483 136 21 MERTOĞLU MERTOĞLU NNS 12483 136 22 AND and CC 12483 136 23 GENÇ genç VBP 12483 136 24 9 9 CD 12483 136 25 news news NN 12483 136 26 item item NN 12483 136 27 , , , 12483 136 28 they -PRON- PRP 12483 136 29 tag tag VBP 12483 136 30 them -PRON- PRP 12483 136 31 as as IN 12483 136 32 fake fake JJ 12483 136 33 , , , 12483 136 34 true true JJ 12483 136 35 , , , 12483 136 36 or or CC 12483 136 37 uncertain uncertain JJ 12483 136 38 . . . 12483 137 1 We -PRON- PRP 12483 137 2 used use VBD 12483 137 3 their -PRON- PRP$ 12483 137 4 results result NNS 12483 137 5 to to TO 12483 137 6 automatically automatically RB 12483 137 7 download download VB 12483 137 8 and and CC 12483 137 9 label label VB 12483 137 10 each each DT 12483 137 11 news news NN 12483 137 12 text text NN 12483 137 13 . . . 12483 138 1 • • NNP 12483 138 2 Lastly lastly RB 12483 138 3 , , , 12483 138 4 our -PRON- PRP$ 12483 138 5 team team NN 12483 138 6 collected collect VBD 12483 138 7 manually manually RB 12483 138 8 curated curate VBN 12483 138 9 and and CC 12483 138 10 verified verify VBN 12483 138 11 fake fake JJ 12483 138 12 and and CC 12483 138 13 valid valid JJ 12483 138 14 news news NN 12483 138 15 obtained obtain VBN 12483 138 16 from from IN 12483 138 17 various various JJ 12483 138 18 online online JJ 12483 138 19 sources source NNS 12483 138 20 and and CC 12483 138 21 named name VBD 12483 138 22 it -PRON- PRP 12483 138 23 as as IN 12483 138 24 MVN MVN NNP 12483 138 25 ( ( -LRB- 12483 138 26 Manually Manually NNP 12483 138 27 Verified Verified NNP 12483 138 28 News News NNP 12483 138 29 ) ) -RRB- 12483 138 30 . . . 12483 139 1 This this DT 12483 139 2 set set NN 12483 139 3 includes include VBZ 12483 139 4 fake fake JJ 12483 139 5 and and CC 12483 139 6 valid valid JJ 12483 139 7 news news NN 12483 139 8 that that WDT 12483 139 9 we -PRON- PRP 12483 139 10 have have VBP 12483 139 11 manually manually RB 12483 139 12 accumulated accumulate VBN 12483 139 13 in in IN 12483 139 14 time time NN 12483 139 15 during during IN 12483 139 16 our -PRON- PRP$ 12483 139 17 studies study NNS 12483 139 18 and and CC 12483 139 19 that that WDT 12483 139 20 were be VBD 12483 139 21 not not RB 12483 139 22 overlapping overlap VBG 12483 139 23 with with IN 12483 139 24 the the DT 12483 139 25 news news NN 12483 139 26 obtained obtain VBN 12483 139 27 from from IN 12483 139 28 GDELT GDELT NNP 12483 139 29 and and CC 12483 139 30 teyit.org teyit.org NNP 12483 139 31 sources source NNS 12483 139 32 . . . 12483 140 1 We -PRON- PRP 12483 140 2 named name VBD 12483 140 3 our -PRON- PRP$ 12483 140 4 dataset dataset JJ 12483 140 5 TRFN TRFN NNP 12483 140 6 . . . 12483 141 1 In in IN 12483 141 2 Phase Phase NNP 12483 141 3 2 2 CD 12483 141 4 , , , 12483 141 5 the the DT 12483 141 6 data data NN 12483 141 7 is be VBZ 12483 141 8 very very RB 12483 141 9 similar similar JJ 12483 141 10 to to IN 12483 141 11 the the DT 12483 141 12 one one NN 12483 141 13 we -PRON- PRP 12483 141 14 used use VBD 12483 141 15 in in IN 12483 141 16 Phase Phase NNP 12483 141 17 1 1 CD 12483 141 18 . . . 12483 142 1 However however RB 12483 142 2 , , , 12483 142 3 to to TO 12483 142 4 see see VB 12483 142 5 the the DT 12483 142 6 effectiveness effectiveness NN 12483 142 7 of of IN 12483 142 8 model model NN 12483 142 9 , , , 12483 142 10 we -PRON- PRP 12483 142 11 made make VBD 12483 142 12 modifications modification NNS 12483 142 13 to to TO 12483 142 14 exclude exclude VB 12483 142 15 old old JJ 12483 142 16 news news NN 12483 142 17 before before IN 12483 142 18 2017 2017 CD 12483 142 19 and and CC 12483 142 20 added add VBD 12483 142 21 new new JJ 12483 142 22 items item NNS 12483 142 23 from from IN 12483 142 24 2019 2019 CD 12483 142 25 . . . 12483 143 1 The the DT 12483 143 2 news news NN 12483 143 3 in in IN 12483 143 4 our -PRON- PRP$ 12483 143 5 dataset dataset NN 12483 143 6 span span VBP 12483 143 7 a a DT 12483 143 8 time time NN 12483 143 9 frame frame NN 12483 143 10 between between IN 12483 143 11 2017 2017 CD 12483 143 12 – – : 12483 143 13 2019 2019 CD 12483 143 14 and and CC 12483 143 15 are be VBP 12483 143 16 uniformly uniformly RB 12483 143 17 distributed distribute VBN 12483 143 18 . . . 12483 144 1 Table table NN 12483 144 2 1 1 CD 12483 144 3 outlines outline VBZ 12483 144 4 the the DT 12483 144 5 dataset dataset JJ 12483 144 6 statistics statistic NNS 12483 144 7 , , , 12483 144 8 namely namely RB 12483 144 9 where where WRB 12483 144 10 the the DT 12483 144 11 news news NN 12483 144 12 text text NN 12483 144 13 comes come VBZ 12483 144 14 from from IN 12483 144 15 , , , 12483 144 16 its -PRON- PRP$ 12483 144 17 class class NN 12483 144 18 ( ( -LRB- 12483 144 19 fake fake JJ 12483 144 20 or or CC 12483 144 21 valid valid JJ 12483 144 22 ) ) -RRB- 12483 144 23 , , , 12483 144 24 the the DT 12483 144 25 amount amount NN 12483 144 26 of of IN 12483 144 27 distinct distinct JJ 12483 144 28 texts text NNS 12483 144 29 and and CC 12483 144 30 the the DT 12483 144 31 corresponding corresponding JJ 12483 144 32 data datum NNS 12483 144 33 collection collection NN 12483 144 34 method method NN 12483 144 35 . . . 12483 145 1 It -PRON- PRP 12483 145 2 can can MD 12483 145 3 be be VB 12483 145 4 seen see VBN 12483 145 5 from from IN 12483 145 6 the the DT 12483 145 7 table table NN 12483 145 8 that that IN 12483 145 9 most most JJS 12483 145 10 of of IN 12483 145 11 our -PRON- PRP$ 12483 145 12 valid valid JJ 12483 145 13 news news NN 12483 145 14 come come VBP 12483 145 15 from from IN 12483 145 16 the the DT 12483 145 17 GDELT GDELT NNP 12483 145 18 source source NN 12483 145 19 , , , 12483 145 20 whereas whereas IN 12483 145 21 teyit.org teyit.org NNP 12483 145 22 , , , 12483 145 23 a a DT 12483 145 24 fact fact NN 12483 145 25 - - HYPH 12483 145 26 checking check VBG 12483 145 27 organization organization NN 12483 145 28 , , , 12483 145 29 contributes contribute VBZ 12483 145 30 only only RB 12483 145 31 fake fake JJ 12483 145 32 news news NN 12483 145 33 . . . 12483 146 1 Table table NN 12483 146 2 1 1 CD 12483 146 3 . . . 12483 147 1 TRFN TRFN NNP 12483 147 2 Dataset Dataset NNP 12483 147 3 Summary Summary NNP 12483 147 4 after after IN 12483 147 5 cleaning clean VBG 12483 147 6 and and CC 12483 147 7 duplicate duplicate VBP 12483 147 8 removal removal NN 12483 147 9 . . . 12483 148 1 Dataset Dataset NNP 12483 148 2 Class Class NNP 12483 148 3 Size Size NNP 12483 148 4 of of IN 12483 148 5 Processed Processed NNP 12483 148 6 Data Data NNP 12483 148 7 Collection Collection NNP 12483 148 8 Method Method NNP 12483 148 9 GDELT GDELT NNP 12483 148 10 NON NON NNP 12483 148 11 - - JJ 12483 148 12 FAKE fake JJ 12483 148 13 82708 82708 CD 12483 148 14 Automated Automated NNP 12483 148 15 Teyit.org Teyit.org NNP 12483 148 16 FAKE FAKE NNP 12483 148 17 1026 1026 CD 12483 148 18 MVN MVN NNP 12483 148 19 NON NON NNP 12483 148 20 - - JJ 12483 148 21 FAKE fake JJ 12483 148 22 1049 1049 CD 12483 148 23 Manual Manual NNP 12483 148 24 FAKE fake NN 12483 148 25 400 400 CD 12483 148 26 All all DT 12483 148 27 news news NN 12483 148 28 items item NNS 12483 148 29 were be VBD 12483 148 30 processed process VBN 12483 148 31 through through IN 12483 148 32 Zemberek Zemberek NNP 12483 148 33 ( ( -LRB- 12483 148 34 http://code.google.com/p/zemberek http://code.google.com/p/zemberek NNP 12483 148 35 ) ) -RRB- 12483 148 36 , , , 12483 148 37 the the DT 12483 148 38 Turkish Turkish NNP 12483 148 39 NLP NLP NNP 12483 148 40 engine engine NN 12483 148 41 for for IN 12483 148 42 extracting extract VBG 12483 148 43 different different JJ 12483 148 44 morphological morphological JJ 12483 148 45 properties property NNS 12483 148 46 of of IN 12483 148 47 words word NNS 12483 148 48 within within IN 12483 148 49 texts text NNS 12483 148 50 . . . 12483 149 1 After after IN 12483 149 2 this this DT 12483 149 3 processing processing NN 12483 149 4 phase phase NN 12483 149 5 , , , 12483 149 6 all all DT 12483 149 7 obtained obtain VBN 12483 149 8 features feature NNS 12483 149 9 were be VBD 12483 149 10 converted convert VBN 12483 149 11 into into IN 12483 149 12 tabular tabular JJ 12483 149 13 format format NN 12483 149 14 and and CC 12483 149 15 made make VBN 12483 149 16 available available JJ 12483 149 17 for for IN 12483 149 18 future future JJ 12483 149 19 studies study NNS 12483 149 20 . . . 12483 150 1 This this DT 12483 150 2 dataset dataset NN 12483 150 3 is be VBZ 12483 150 4 now now RB 12483 150 5 available available JJ 12483 150 6 for for IN 12483 150 7 scholarly scholarly JJ 12483 150 8 studies study NNS 12483 150 9 upon upon IN 12483 150 10 request request NN 12483 150 11 . . . 12483 151 1 In in IN 12483 151 2 a a DT 12483 151 3 study study NN 12483 151 4 of of IN 12483 151 5 this this DT 12483 151 6 nature nature NN 12483 151 7 , , , 12483 151 8 the the DT 12483 151 9 verifiability verifiability NN 12483 151 10 of of IN 12483 151 11 the the DT 12483 151 12 data datum NNS 12483 151 13 used use VBN 12483 151 14 is be VBZ 12483 151 15 important important JJ 12483 151 16 . . . 12483 152 1 As as IN 12483 152 2 we -PRON- PRP 12483 152 3 have have VBP 12483 152 4 already already RB 12483 152 5 mentioned mention VBN 12483 152 6 , , , 12483 152 7 most most JJS 12483 152 8 of of IN 12483 152 9 the the DT 12483 152 10 data datum NNS 12483 152 11 we -PRON- PRP 12483 152 12 used use VBD 12483 152 13 comes come VBZ 12483 152 14 from from IN 12483 152 15 verified verified JJ 12483 152 16 sources source NNS 12483 152 17 such such JJ 12483 152 18 as as IN 12483 152 19 mainstream mainstream JJ 12483 152 20 news news NN 12483 152 21 agencies agency NNS 12483 152 22 accessed access VBN 12483 152 23 through through IN 12483 152 24 GDELT GDELT NNP 12483 152 25 and and CC 12483 152 26 teyit.org teyit.org CD 12483 152 27 archives archive NNS 12483 152 28 which which WDT 12483 152 29 are be VBP 12483 152 30 verified verify VBN 12483 152 31 by by IN 12483 152 32 teyit.org teyit.org NN 12483 152 33 staff staff NN 12483 152 34 . . . 12483 153 1 All all DT 12483 153 2 data datum NNS 12483 153 3 used use VBN 12483 153 4 in in IN 12483 153 5 training train VBG 12483 153 6 the the DT 12483 153 7 mathematical mathematical JJ 12483 153 8 models model NNS 12483 153 9 which which WDT 12483 153 10 are be VBP 12483 153 11 to to TO 12483 153 12 be be VB 12483 153 13 explained explain VBN 12483 153 14 in in IN 12483 153 15 the the DT 12483 153 16 rest rest NN 12483 153 17 of of IN 12483 153 18 the the DT 12483 153 19 paper paper NN 12483 153 20 are be VBP 12483 153 21 either either RB 12483 153 22 directly directly RB 12483 153 23 or or CC 12483 153 24 indirectly indirectly RB 12483 153 25 verified verify VBD 12483 153 26 . . . 12483 154 1 Another another DT 12483 154 2 important important JJ 12483 154 3 issue issue NN 12483 154 4 was be VBD 12483 154 5 generalizability generalizability NN 12483 154 6 of of IN 12483 154 7 the the DT 12483 154 8 dataset dataset NN 12483 154 9 , , , 12483 154 10 which which WDT 12483 154 11 determines determine VBZ 12483 154 12 whether whether IN 12483 154 13 the the DT 12483 154 14 results result NNS 12483 154 15 of of IN 12483 154 16 the the DT 12483 154 17 study study NN 12483 154 18 are be VBP 12483 154 19 only only RB 12483 154 20 applicable applicable JJ 12483 154 21 to to IN 12483 154 22 specific specific JJ 12483 154 23 domains domain NNS 12483 154 24 or or CC 12483 154 25 to to IN 12483 154 26 all all DT 12483 154 27 available available JJ 12483 154 28 domains domain NNS 12483 154 29 . . . 12483 155 1 Although although IN 12483 155 2 focusing focus VBG 12483 155 3 on on IN 12483 155 4 a a DT 12483 155 5 specific specific JJ 12483 155 6 news news NN 12483 155 7 domain domain NN 12483 155 8 would would MD 12483 155 9 clearly clearly RB 12483 155 10 improve improve VB 12483 155 11 our -PRON- PRP$ 12483 155 12 accuracies accuracy NNS 12483 155 13 , , , 12483 155 14 we -PRON- PRP 12483 155 15 preferred prefer VBD 12483 155 16 to to TO 12483 155 17 work work VB 12483 155 18 in in IN 12483 155 19 the the DT 12483 155 20 general general JJ 12483 155 21 domain domain NN 12483 155 22 and and CC 12483 155 23 included include VBD 12483 155 24 news news NN 12483 155 25 from from IN 12483 155 26 all all DT 12483 155 27 specific specific JJ 12483 155 28 domains domain NNS 12483 155 29 . . . 12483 156 1 The the DT 12483 156 2 distribution distribution NN 12483 156 3 of of IN 12483 156 4 domains domain NNS 12483 156 5 in in IN 12483 156 6 our -PRON- PRP$ 12483 156 7 dataset dataset NN 12483 156 8 is be VBZ 12483 156 9 visualized visualize VBN 12483 156 10 in in IN 12483 156 11 figure figure NN 12483 156 12 4 4 CD 12483 156 13 . . . 12483 157 1 This this DT 12483 157 2 distribution distribution NN 12483 157 3 closely closely RB 12483 157 4 matches match VBZ 12483 157 5 the the DT 12483 157 6 distribution distribution NN 12483 157 7 one one CD 12483 157 8 would would MD 12483 157 9 experience experience VB 12483 157 10 reading read VBG 12483 157 11 daily daily JJ 12483 157 12 news news NN 12483 157 13 in in IN 12483 157 14 Turkey Turkey NNP 12483 157 15 . . . 12483 158 1 Hence hence RB 12483 158 2 , , , 12483 158 3 we -PRON- PRP 12483 158 4 have have VBP 12483 158 5 no no DT 12483 158 6 domain domain NN 12483 158 7 specific specific JJ 12483 158 8 bias bias NN 12483 158 9 in in IN 12483 158 10 our -PRON- PRP$ 12483 158 11 training training NN 12483 158 12 dataset dataset NN 12483 158 13 . . . 12483 159 1 http://code.google.com/p/zemberek http://code.google.com/p/zemberek ADD 12483 159 2 INFORMATION INFORMATION NNP 12483 159 3 TECHNOLOGY technology NN 12483 159 4 AND and CC 12483 159 5 LIBRARIES library NNS 12483 159 6 DECEMBER DECEMBER NNP 12483 159 7 2020 2020 CD 12483 159 8 AUTOMATED AUTOMATED NNP 12483 159 9 FAKE FAKE NNP 12483 159 10 NEWS NEWS NNP 12483 159 11 DETECTION detection VB 12483 159 12 IN in IN 12483 159 13 THE the DT 12483 159 14 AGE AGE NNP 12483 159 15 OF of IN 12483 159 16 DIGITAL DIGITAL NNP 12483 159 17 LIBRARIES library NNS 12483 159 18 | | NNP 12483 159 19 MERTOĞLU MERTOĞLU NNS 12483 159 20 AND and CC 12483 159 21 GENÇ genç VBP 12483 159 22 10 10 CD 12483 159 23 Figure figure NN 12483 159 24 4 4 CD 12483 159 25 . . . 12483 160 1 The the DT 12483 160 2 distribution distribution NN 12483 160 3 of of IN 12483 160 4 domains domain NNS 12483 160 5 in in IN 12483 160 6 the the DT 12483 160 7 dataset dataset NN 12483 160 8 . . . 12483 161 1 ( ( -LRB- 12483 161 2 SciTechEnvWetNatLife scitechenvwetnatlife NN 12483 161 3 = = SYM 12483 161 4 Science Science NNP 12483 161 5 , , , 12483 161 6 Technology Technology NNP 12483 161 7 , , , 12483 161 8 Environment Environment NNP 12483 161 9 , , , 12483 161 10 Weather Weather NNP 12483 161 11 , , , 12483 161 12 Nature Nature NNP 12483 161 13 , , , 12483 161 14 Life life NN 12483 161 15 . . . 12483 162 1 EduCultureArtTourism educulturearttourism NN 12483 162 2 = = SYM 12483 162 3 Education Education NNP 12483 162 4 , , , 12483 162 5 Culture Culture NNP 12483 162 6 , , , 12483 162 7 Art Art NNP 12483 162 8 , , , 12483 162 9 Tourism Tourism NNP 12483 162 10 . . . 12483 162 11 ) ) -RRB- 12483 163 1 Moreover moreover RB 12483 163 2 , , , 12483 163 3 we -PRON- PRP 12483 163 4 obtained obtain VBD 12483 163 5 highly highly RB 12483 163 6 correlated correlate VBN 12483 163 7 evidence evidence NN 12483 163 8 showing show VBG 12483 163 9 syntactic syntactic JJ 12483 163 10 similarities similarity NNS 12483 163 11 with with IN 12483 163 12 the the DT 12483 163 13 other other JJ 12483 163 14 NLP NLP NNP 12483 163 15 studies study NNS 12483 163 16 in in IN 12483 163 17 Turkish Turkish NNP 12483 163 18 during during IN 12483 163 19 the the DT 12483 163 20 exploratory exploratory JJ 12483 163 21 data data NN 12483 163 22 analysis analysis NN 12483 163 23 . . . 12483 164 1 For for IN 12483 164 2 example example NN 12483 164 3 , , , 12483 164 4 the the DT 12483 164 5 results result NNS 12483 164 6 of of IN 12483 164 7 a a DT 12483 164 8 study study NN 12483 164 9 by by IN 12483 164 10 Zemberek Zemberek NNP 12483 164 11 developers developer NNS 12483 164 12 ( ( -LRB- 12483 164 13 http://zembereknlp.blogspot.com/2006/11/kelime-istatistikleri.html http://zembereknlp.blogspot.com/2006/11/kelime-istatistikleri.html NN 12483 164 14 ) ) -RRB- 12483 164 15 to to TO 12483 164 16 find find VB 12483 164 17 the the DT 12483 164 18 most most RBS 12483 164 19 common common JJ 12483 164 20 words word NNS 12483 164 21 in in IN 12483 164 22 Turkish turkish JJ 12483 164 23 experimented experiment VBD 12483 164 24 with with IN 12483 164 25 over over IN 12483 164 26 five five CD 12483 164 27 million million CD 12483 164 28 words word NNS 12483 164 29 is be VBZ 12483 164 30 compatible compatible JJ 12483 164 31 with with IN 12483 164 32 most most JJS 12483 164 33 common common JJ 12483 164 34 words word NNS 12483 164 35 in in IN 12483 164 36 our -PRON- PRP$ 12483 164 37 corpus corpus NN 12483 164 38 . . . 12483 165 1 This this DT 12483 165 2 evidence evidence NN 12483 165 3 can can MD 12483 165 4 be be VB 12483 165 5 attributed attribute VBN 12483 165 6 to to IN 12483 165 7 representability representability NN 12483 165 8 of of IN 12483 165 9 our -PRON- PRP$ 12483 165 10 dataset dataset NN 12483 165 11 . . . 12483 166 1 The the DT 12483 166 2 last last JJ 12483 166 3 issue issue NN 12483 166 4 worth worth JJ 12483 166 5 discussing discuss VBG 12483 166 6 is be VBZ 12483 166 7 the the DT 12483 166 8 imbalanced imbalanced JJ 12483 166 9 nature nature NN 12483 166 10 of of IN 12483 166 11 the the DT 12483 166 12 dataset dataset NN 12483 166 13 . . . 12483 167 1 An an DT 12483 167 2 imbalanced imbalanced JJ 12483 167 3 dataset dataset NN 12483 167 4 occurs occur VBZ 12483 167 5 in in IN 12483 167 6 a a DT 12483 167 7 binary binary JJ 12483 167 8 classification classification NN 12483 167 9 study study NN 12483 167 10 when when WRB 12483 167 11 the the DT 12483 167 12 frequency frequency NN 12483 167 13 of of IN 12483 167 14 one one CD 12483 167 15 of of IN 12483 167 16 the the DT 12483 167 17 classes class NNS 12483 167 18 dominates dominate VBZ 12483 167 19 the the DT 12483 167 20 frequency frequency NN 12483 167 21 of of IN 12483 167 22 the the DT 12483 167 23 other other JJ 12483 167 24 class class NN 12483 167 25 in in IN 12483 167 26 the the DT 12483 167 27 dataset dataset NN 12483 167 28 . . . 12483 168 1 In in IN 12483 168 2 our -PRON- PRP$ 12483 168 3 dataset dataset NN 12483 168 4 , , , 12483 168 5 the the DT 12483 168 6 amount amount NN 12483 168 7 of of IN 12483 168 8 fake fake JJ 12483 168 9 news news NN 12483 168 10 is be VBZ 12483 168 11 highly highly RB 12483 168 12 surpassed surpass VBN 12483 168 13 by by IN 12483 168 14 the the DT 12483 168 15 amount amount NN 12483 168 16 of of IN 12483 168 17 valid valid JJ 12483 168 18 news news NN 12483 168 19 . . . 12483 169 1 This this DT 12483 169 2 generally generally RB 12483 169 3 results result VBZ 12483 169 4 in in IN 12483 169 5 difficulties difficulty NNS 12483 169 6 in in IN 12483 169 7 applying apply VBG 12483 169 8 conventional conventional JJ 12483 169 9 machine machine NN 12483 169 10 learning learning NN 12483 169 11 methods method NNS 12483 169 12 to to IN 12483 169 13 the the DT 12483 169 14 dataset dataset NN 12483 169 15 . . . 12483 170 1 However however RB 12483 170 2 , , , 12483 170 3 it -PRON- PRP 12483 170 4 is be VBZ 12483 170 5 a a DT 12483 170 6 frequently frequently RB 12483 170 7 observed observe VBN 12483 170 8 phenomenon phenomenon NN 12483 170 9 due due IN 12483 170 10 to to IN 12483 170 11 the the DT 12483 170 12 disparity disparity NN 12483 170 13 of of IN 12483 170 14 variable variable JJ 12483 170 15 classes class NNS 12483 170 16 in in IN 12483 170 17 these these DT 12483 170 18 kinds kind NNS 12483 170 19 of of IN 12483 170 20 problems problem NNS 12483 170 21 in in IN 12483 170 22 real real JJ 12483 170 23 world world NN 12483 170 24 . . . 12483 171 1 To to TO 12483 171 2 avoid avoid VB 12483 171 3 potential potential JJ 12483 171 4 problems problem NNS 12483 171 5 due due JJ 12483 171 6 to to IN 12483 171 7 the the DT 12483 171 8 imbalanced imbalanced JJ 12483 171 9 nature nature NN 12483 171 10 of of IN 12483 171 11 the the DT 12483 171 12 dataset dataset NN 12483 171 13 , , , 12483 171 14 we -PRON- PRP 12483 171 15 used use VBD 12483 171 16 SMOTE SMOTE NNP 12483 171 17 ( ( -LRB- 12483 171 18 Synthetic synthetic JJ 12483 171 19 http://zembereknlp.blogspot.com/2006/11/kelime-istatistikleri.html http://zembereknlp.blogspot.com/2006/11/kelime-istatistikleri.html NN 12483 171 20 INFORMATION INFORMATION VBD 12483 171 21 TECHNOLOGY technology NN 12483 171 22 AND and CC 12483 171 23 LIBRARIES library NNS 12483 171 24 DECEMBER DECEMBER NNP 12483 171 25 2020 2020 CD 12483 171 26 AUTOMATED AUTOMATED NNP 12483 171 27 FAKE FAKE NNP 12483 171 28 NEWS NEWS NNP 12483 171 29 DETECTION detection VB 12483 171 30 IN in IN 12483 171 31 THE the DT 12483 171 32 AGE AGE NNP 12483 171 33 OF of IN 12483 171 34 DIGITAL DIGITAL NNP 12483 171 35 LIBRARIES library NNS 12483 171 36 | | NNP 12483 171 37 MERTOĞLU MERTOĞLU NNS 12483 171 38 AND and CC 12483 171 39 GENÇ genç VBP 12483 171 40 11 11 CD 12483 171 41 Minority Minority NNP 12483 171 42 Over over IN 12483 171 43 - - HYPH 12483 171 44 sampling sampling NN 12483 171 45 Technique Technique NNP 12483 171 46 ) ) -RRB- 12483 171 47 which which WDT 12483 171 48 is be VBZ 12483 171 49 an an DT 12483 171 50 over over IN 12483 171 51 - - HYPH 12483 171 52 sampling sampling NN 12483 171 53 method.30 method.30 NNP 12483 171 54 It -PRON- PRP 12483 171 55 creates create VBZ 12483 171 56 synthetic synthetic JJ 12483 171 57 samples sample NNS 12483 171 58 of of IN 12483 171 59 the the DT 12483 171 60 minority minority NN 12483 171 61 class class NN 12483 171 62 that that WDT 12483 171 63 are be VBP 12483 171 64 relatively relatively RB 12483 171 65 close close JJ 12483 171 66 in in IN 12483 171 67 the the DT 12483 171 68 feature feature NN 12483 171 69 space space NN 12483 171 70 to to IN 12483 171 71 the the DT 12483 171 72 existing exist VBG 12483 171 73 observations observation NNS 12483 171 74 of of IN 12483 171 75 the the DT 12483 171 76 minority minority NN 12483 171 77 class class NN 12483 171 78 . . . 12483 172 1 Features feature NNS 12483 172 2 In in IN 12483 172 3 this this DT 12483 172 4 study study NN 12483 172 5 , , , 12483 172 6 we -PRON- PRP 12483 172 7 discarded discard VBD 12483 172 8 some some DT 12483 172 9 features feature NNS 12483 172 10 because because IN 12483 172 11 of of IN 12483 172 12 their -PRON- PRP$ 12483 172 13 relatively relatively RB 12483 172 14 low low JJ 12483 172 15 impact impact NN 12483 172 16 on on IN 12483 172 17 overall overall JJ 12483 172 18 performance performance NN 12483 172 19 during during IN 12483 172 20 the the DT 12483 172 21 exploratory exploratory NNP 12483 172 22 data data NN 12483 172 23 analysis analysis NN 12483 172 24 and and CC 12483 172 25 subsequently subsequently RB 12483 172 26 in in IN 12483 172 27 the the DT 12483 172 28 training training NN 12483 172 29 phase phase NN 12483 172 30 . . . 12483 173 1 The the DT 12483 173 2 most most RBS 12483 173 3 effective effective JJ 12483 173 4 features feature NNS 12483 173 5 we -PRON- PRP 12483 173 6 decided decide VBD 12483 173 7 on on IN 12483 173 8 are be VBP 12483 173 9 shown show VBN 12483 173 10 in in IN 12483 173 11 table table NN 12483 173 12 2 2 CD 12483 173 13 . . . 12483 174 1 Table table NN 12483 174 2 2 2 CD 12483 174 3 . . . 12483 175 1 Main main JJ 12483 175 2 Features Features NNP 12483 175 3 Features Features NNPS 12483 175 4 Group Group NNP 12483 175 5 Definition Definition NNP 12483 175 6 nRootScore nRootScore NNP 12483 175 7 Language Language NNP 12483 175 8 Model Model NNP 12483 175 9 Features feature VBZ 12483 175 10 The the DT 12483 175 11 news news NN 12483 175 12 score score NN 12483 175 13 calculated calculate VBN 12483 175 14 according accord VBG 12483 175 15 to to IN 12483 175 16 the the DT 12483 175 17 Root Root NNP 12483 175 18 Model Model NNP 12483 175 19 nRawScore nRawScore NNP 12483 175 20 The the DT 12483 175 21 news news NN 12483 175 22 score score NN 12483 175 23 calculated calculate VBN 12483 175 24 according accord VBG 12483 175 25 to to IN 12483 175 26 the the DT 12483 175 27 Raw Raw NNP 12483 175 28 Model Model NNP 12483 175 29 SpellErrorScore spellerrorscore NN 12483 175 30 Extracted Extracted NNP 12483 175 31 Features Features NNP 12483 175 32 Spell Spell NNP 12483 175 33 errors error NNS 12483 175 34 per per IN 12483 175 35 sentences sentence NNS 12483 175 36 ComplexityScore ComplexityScore NNP 12483 175 37 The the DT 12483 175 38 score score NN 12483 175 39 of of IN 12483 175 40 the the DT 12483 175 41 complexity complexity NN 12483 175 42 / / SYM 12483 175 43 readibility readibility NN 12483 175 44 of of IN 12483 175 45 the the DT 12483 175 46 news news NN 12483 175 47 Source source NN 12483 175 48 Labels Labels NNP 12483 175 49 The the DT 12483 175 50 URL URL NNP 12483 175 51 or or CC 12483 175 52 identifier identifier NN 12483 175 53 of of IN 12483 175 54 the the DT 12483 175 55 news news NN 12483 175 56 MainCategory MainCategory NNP 12483 175 57 The the DT 12483 175 58 category category NN 12483 175 59 of of IN 12483 175 60 the the DT 12483 175 61 news news NN 12483 175 62 NewsSite NewsSite NNP 12483 175 63 The the DT 12483 175 64 unique unique JJ 12483 175 65 address address NN 12483 175 66 of of IN 12483 175 67 the the DT 12483 175 68 news news NN 12483 175 69 The the DT 12483 175 70 language language NN 12483 175 71 model model NN 12483 175 72 features feature VBZ 12483 175 73 nRootScore nrootscore CD 12483 175 74 and and CC 12483 175 75 nRawScore nRawScore NNP 12483 175 76 are be VBP 12483 175 77 features feature NNS 12483 175 78 that that WDT 12483 175 79 we -PRON- PRP 12483 175 80 have have VBP 12483 175 81 borrowed borrow VBN 12483 175 82 from from IN 12483 175 83 our -PRON- PRP$ 12483 175 84 earlier early JJR 12483 175 85 study study NN 12483 175 86 on on IN 12483 175 87 fake fake JJ 12483 175 88 news news NN 12483 175 89 detection.31 detection.31 NNP 12483 175 90 In in IN 12483 175 91 that that DT 12483 175 92 study study NN 12483 175 93 , , , 12483 175 94 we -PRON- PRP 12483 175 95 focused focus VBD 12483 175 96 on on IN 12483 175 97 constructing construct VBG 12483 175 98 a a DT 12483 175 99 fake fake JJ 12483 175 100 news news NN 12483 175 101 dictionary dictionary NNP 12483 175 102 / / SYM 12483 175 103 lexicon lexicon NNS 12483 175 104 based base VBN 12483 175 105 on on IN 12483 175 106 different different JJ 12483 175 107 morphological morphological JJ 12483 175 108 segments segment NNS 12483 175 109 of of IN 12483 175 110 the the DT 12483 175 111 words word NNS 12483 175 112 used use VBN 12483 175 113 in in IN 12483 175 114 news news NN 12483 175 115 texts text NNS 12483 175 116 . . . 12483 176 1 These these DT 12483 176 2 two two CD 12483 176 3 scores score NNS 12483 176 4 were be VBD 12483 176 5 found find VBN 12483 176 6 to to TO 12483 176 7 be be VB 12483 176 8 the the DT 12483 176 9 most most RBS 12483 176 10 successful successful JJ 12483 176 11 ones one NNS 12483 176 12 in in IN 12483 176 13 determining determine VBG 12483 176 14 the the DT 12483 176 15 fakeness fakeness NN 12483 176 16 / / SYM 12483 176 17 validity validity NN 12483 176 18 of of IN 12483 176 19 a a DT 12483 176 20 news news NN 12483 176 21 text text NN 12483 176 22 , , , 12483 176 23 one one CD 12483 176 24 considering consider VBG 12483 176 25 the the DT 12483 176 26 raw raw JJ 12483 176 27 form form NN 12483 176 28 of of IN 12483 176 29 the the DT 12483 176 30 words word NNS 12483 176 31 , , , 12483 176 32 the the DT 12483 176 33 other other JJ 12483 176 34 considering consider VBG 12483 176 35 the the DT 12483 176 36 root root NN 12483 176 37 form form NN 12483 176 38 . . . 12483 177 1 The the DT 12483 177 2 extracted extracted JJ 12483 177 3 features feature NNS 12483 177 4 are be VBP 12483 177 5 ComplexityScore ComplexityScore NNP 12483 177 6 and and CC 12483 177 7 SpellErrorScore SpellErrorScore : 12483 177 8 . . . 12483 178 1 ComplexityScore ComplexityScore NNP 12483 178 2 basically basically RB 12483 178 3 represents represent VBZ 12483 178 4 the the DT 12483 178 5 readability readability NN 12483 178 6 of of IN 12483 178 7 the the DT 12483 178 8 text text NN 12483 178 9 . . . 12483 179 1 Studies study NNS 12483 179 2 for for IN 12483 179 3 determining determine VBG 12483 179 4 a a DT 12483 179 5 good good JJ 12483 179 6 readability readability NN 12483 179 7 metric metric JJ 12483 179 8 exist exist VBP 12483 179 9 for for IN 12483 179 10 the the DT 12483 179 11 Turkish turkish JJ 12483 179 12 language.32 language.32 NNP 12483 179 13 We -PRON- PRP 12483 179 14 used use VBD 12483 179 15 a a DT 12483 179 16 modified modify VBN 12483 179 17 version version NN 12483 179 18 of of IN 12483 179 19 the the DT 12483 179 20 Gunnig Gunnig NNP 12483 179 21 - - HYPH 12483 179 22 Fog Fog NNP 12483 179 23 metric metric NN 12483 179 24 , , , 12483 179 25 which which WDT 12483 179 26 is be VBZ 12483 179 27 based base VBN 12483 179 28 on on IN 12483 179 29 word word NN 12483 179 30 length length NN 12483 179 31 and and CC 12483 179 32 sentence sentence NN 12483 179 33 length.33 length.33 NNP 12483 179 34 Since since IN 12483 179 35 Turkish Turkish NNP 12483 179 36 is be VBZ 12483 179 37 an an DT 12483 179 38 agglutinative agglutinative JJ 12483 179 39 language language NN 12483 179 40 , , , 12483 179 41 we -PRON- PRP 12483 179 42 used use VBD 12483 179 43 word word NN 12483 179 44 length length NN 12483 179 45 instead instead RB 12483 179 46 of of IN 12483 179 47 using use VBG 12483 179 48 the the DT 12483 179 49 syllable syllable NN 12483 179 50 count count NN 12483 179 51 . . . 12483 180 1 We -PRON- PRP 12483 180 2 also also RB 12483 180 3 made make VBD 12483 180 4 some some DT 12483 180 5 modifications modification NNS 12483 180 6 to to TO 12483 180 7 normalize normalize VB 12483 180 8 the the DT 12483 180 9 scores score NNS 12483 180 10 . . . 12483 181 1 The the DT 12483 181 2 average average JJ 12483 181 3 number number NN 12483 181 4 of of IN 12483 181 5 syllables syllable NNS 12483 181 6 per per IN 12483 181 7 word word NN 12483 181 8 syllable syllable NN 12483 181 9 in in IN 12483 181 10 Turkish Turkish NNP 12483 181 11 is be VBZ 12483 181 12 2.6 2.6 CD 12483 181 13 , , , 12483 181 14 so so RB 12483 181 15 we -PRON- PRP 12483 181 16 defined define VBD 12483 181 17 a a DT 12483 181 18 word word NN 12483 181 19 as as IN 12483 181 20 a a DT 12483 181 21 long long JJ 12483 181 22 word word NN 12483 181 23 if if IN 12483 181 24 it -PRON- PRP 12483 181 25 has have VBZ 12483 181 26 more more JJR 12483 181 27 than than IN 12483 181 28 9 9 CD 12483 181 29 letters.34 letters.34 NN 12483 181 30 For for IN 12483 181 31 a a DT 12483 181 32 given give VBN 12483 181 33 news news NN 12483 181 34 text text NN 12483 181 35 T T NNP 12483 181 36 , , , 12483 181 37 the the DT 12483 181 38 Complexity Complexity NNP 12483 181 39 Score Score NNP 12483 181 40 ( ( -LRB- 12483 181 41 CS CS NNP 12483 181 42 ) ) -RRB- 12483 181 43 can can MD 12483 181 44 be be VB 12483 181 45 computed compute VBN 12483 181 46 by by IN 12483 181 47 equation equation NN 12483 181 48 1 1 CD 12483 181 49 . . . 12483 182 1 ( ( -LRB- 12483 182 2 1 1 LS 12483 182 3 ) ) -RRB- 12483 182 4 𝑇𝐶𝑆 𝑇𝐶𝑆 NNP 12483 182 5 = = NFP 12483 182 6 ( ( -LRB- 12483 182 7 𝑊𝑜𝑟𝑑𝑐𝑜𝑢𝑛𝑡 𝑊𝑜𝑟𝑑𝑐𝑜𝑢𝑛𝑡 NNP 12483 182 8 𝑆𝑒𝑛𝑡𝑒𝑛𝑐𝑒𝑠𝑐𝑜𝑢𝑛𝑡 𝑆𝑒𝑛𝑡𝑒𝑛𝑐𝑒𝑠𝑐𝑜𝑢𝑛𝑡 NNP 12483 182 9 + + CC 12483 182 10 𝐿𝑜𝑛𝑔𝑊𝑜𝑟𝑑𝑐𝑜𝑢𝑛𝑡∗100 𝐿𝑜𝑛𝑔𝑊𝑜𝑟𝑑𝑐𝑜𝑢𝑛𝑡∗100 VBZ 12483 182 11 𝑊𝑜𝑟𝑑𝑐𝑜𝑢𝑛𝑡 𝑊𝑜𝑟𝑑𝑐𝑜𝑢𝑛𝑡 NNP 12483 182 12 10 10 CD 12483 182 13 ) ) -RRB- 12483 182 14 The the DT 12483 182 15 second second JJ 12483 182 16 Extracted extract VBN 12483 182 17 Feature Feature NNP 12483 182 18 is be VBZ 12483 182 19 SpellErrorScore SpellErrorScore : 12483 182 20 . . . 12483 183 1 We -PRON- PRP 12483 183 2 foresee foresee VBP 12483 183 3 that that IN 12483 183 4 there there EX 12483 183 5 may may MD 12483 183 6 be be VB 12483 183 7 many many JJ 12483 183 8 more more JJR 12483 183 9 errors error NNS 12483 183 10 in in IN 12483 183 11 fake fake JJ 12483 183 12 news news NN 12483 183 13 than than IN 12483 183 14 in in IN 12483 183 15 valid valid JJ 12483 183 16 news news NN 12483 183 17 . . . 12483 184 1 We -PRON- PRP 12483 184 2 calculated calculate VBD 12483 184 3 the the DT 12483 184 4 spell spell NN 12483 184 5 error error NN 12483 184 6 counts count NNS 12483 184 7 making make VBG 12483 184 8 use use NN 12483 184 9 of of IN 12483 184 10 Turkish turkish JJ 12483 184 11 Spellchecker Spellchecker NNP 12483 184 12 class class NN 12483 184 13 of of IN 12483 184 14 Zemberek Zemberek NNP 12483 184 15 . . . 12483 185 1 Due due IN 12483 185 2 to to IN 12483 185 3 the the DT 12483 185 4 text text NN 12483 185 5 length length NN 12483 185 6 of of IN 12483 185 7 news news NN 12483 185 8 varies varie NNS 12483 185 9 , , , 12483 185 10 we -PRON- PRP 12483 185 11 calculate calculate VBP 12483 185 12 the the DT 12483 185 13 ratio ratio NN 12483 185 14 INFORMATION information NN 12483 185 15 TECHNOLOGY technology NN 12483 185 16 AND and CC 12483 185 17 LIBRARIES library NNS 12483 185 18 DECEMBER DECEMBER NNP 12483 185 19 2020 2020 CD 12483 185 20 AUTOMATED AUTOMATED NNP 12483 185 21 FAKE FAKE NNP 12483 185 22 NEWS NEWS NNP 12483 185 23 DETECTION detection VB 12483 185 24 IN in IN 12483 185 25 THE the DT 12483 185 26 AGE AGE NNP 12483 185 27 OF of IN 12483 185 28 DIGITAL DIGITAL NNP 12483 185 29 LIBRARIES library NNS 12483 185 30 | | NNP 12483 185 31 MERTOĞLU MERTOĞLU NNS 12483 185 32 AND and CC 12483 185 33 GENÇ genç VBP 12483 185 34 12 12 CD 12483 185 35 according accord VBG 12483 185 36 to to IN 12483 185 37 the the DT 12483 185 38 sentences sentence NNS 12483 185 39 . . . 12483 186 1 For for IN 12483 186 2 a a DT 12483 186 3 given give VBN 12483 186 4 news news NN 12483 186 5 text text NN 12483 186 6 T T NNP 12483 186 7 , , , 12483 186 8 SE SE NNP 12483 186 9 ( ( -LRB- 12483 186 10 Spell Spell NNP 12483 186 11 Error Error NNP 12483 186 12 Score Score NNP 12483 186 13 ) ) -RRB- 12483 186 14 is be VBZ 12483 186 15 calculated calculate VBN 12483 186 16 as as IN 12483 186 17 shown show VBN 12483 186 18 in in IN 12483 186 19 equation equation NN 12483 186 20 2 2 CD 12483 186 21 . . . 12483 187 1 ( ( -LRB- 12483 187 2 2 2 LS 12483 187 3 ) ) -RRB- 12483 187 4 𝑇𝑆𝐸 𝑇𝑆𝐸 NNP 12483 187 5 = = NFP 12483 187 6 ( ( -LRB- 12483 187 7 𝑆𝑝𝑒𝑙𝑙𝐸𝑟𝑟𝑜𝑟𝐶𝑜𝑢𝑛𝑡 𝑆𝑝𝑒𝑙𝑙𝐸𝑟𝑟𝑜𝑟𝐶𝑜𝑢𝑛𝑡 NFP 12483 187 8 𝑆𝑒𝑛𝑡𝑒𝑛𝑐𝑒𝑠𝐶𝑜𝑢𝑛𝑡 𝑆𝑒𝑛𝑡𝑒𝑛𝑐𝑒𝑠𝐶𝑜𝑢𝑛𝑡 NNP 12483 187 9 ) ) -RRB- 12483 187 10 Finally finally RB 12483 187 11 , , , 12483 187 12 we -PRON- PRP 12483 187 13 included include VBD 12483 187 14 the the DT 12483 187 15 metadata metadata NN 12483 187 16 categories category NNS 12483 187 17 Source Source NNP 12483 187 18 , , , 12483 187 19 MainCategory MainCategory NNP 12483 187 20 , , , 12483 187 21 and and CC 12483 187 22 NewsSite NewsSite NNP 12483 187 23 as as IN 12483 187 24 additional additional JJ 12483 187 25 identifiers identifier NNS 12483 187 26 for for IN 12483 187 27 the the DT 12483 187 28 learning learning NN 12483 187 29 process process NN 12483 187 30 . . . 12483 188 1 Then then RB 12483 188 2 , , , 12483 188 3 we -PRON- PRP 12483 188 4 combined combine VBD 12483 188 5 features feature NNS 12483 188 6 extracted extract VBN 12483 188 7 from from IN 12483 188 8 text text NN 12483 188 9 representation representation NN 12483 188 10 techniques technique NNS 12483 188 11 with with IN 12483 188 12 the the DT 12483 188 13 features feature NNS 12483 188 14 shown show VBN 12483 188 15 in in IN 12483 188 16 table table NN 12483 188 17 2 2 CD 12483 188 18 and and CC 12483 188 19 trained train VBD 12483 188 20 the the DT 12483 188 21 model model NN 12483 188 22 with with IN 12483 188 23 different different JJ 12483 188 24 classifiers classifier NNS 12483 188 25 . . . 12483 189 1 For for IN 12483 189 2 text text NN 12483 189 3 representation representation NN 12483 189 4 , , , 12483 189 5 we -PRON- PRP 12483 189 6 followed follow VBD 12483 189 7 two two CD 12483 189 8 directions direction NNS 12483 189 9 for for IN 12483 189 10 the the DT 12483 189 11 experiments experiment NNS 12483 189 12 . . . 12483 190 1 First first RB 12483 190 2 , , , 12483 190 3 we -PRON- PRP 12483 190 4 converted convert VBD 12483 190 5 text text NN 12483 190 6 into into IN 12483 190 7 structured structured JJ 12483 190 8 features feature NNS 12483 190 9 with with IN 12483 190 10 Bag Bag NNP 12483 190 11 of of IN 12483 190 12 Words Words NNP 12483 190 13 ( ( -LRB- 12483 190 14 BOW BOW NNP 12483 190 15 ) ) -RRB- 12483 190 16 approach approach NN 12483 190 17 in in IN 12483 190 18 which which WDT 12483 190 19 text text NN 12483 190 20 data datum NNS 12483 190 21 is be VBZ 12483 190 22 represented represent VBN 12483 190 23 as as IN 12483 190 24 the the DT 12483 190 25 multiset multiset NN 12483 190 26 of of IN 12483 190 27 its -PRON- PRP$ 12483 190 28 words word NNS 12483 190 29 . . . 12483 191 1 Second second JJ 12483 191 2 , , , 12483 191 3 we -PRON- PRP 12483 191 4 experimented experiment VBD 12483 191 5 with with IN 12483 191 6 N N NNP 12483 191 7 - - HYPH 12483 191 8 grams gram NNS 12483 191 9 which which WDT 12483 191 10 represents represent VBZ 12483 191 11 the the DT 12483 191 12 sequence sequence NN 12483 191 13 of of IN 12483 191 14 n n JJ 12483 191 15 words word NNS 12483 191 16 , , , 12483 191 17 in in IN 12483 191 18 other other JJ 12483 191 19 words word NNS 12483 191 20 splitting split VBG 12483 191 21 text text NN 12483 191 22 into into IN 12483 191 23 chunks chunk NNS 12483 191 24 of of IN 12483 191 25 size size NN 12483 191 26 N n NN 12483 191 27 - - HYPH 12483 191 28 words word NNS 12483 191 29 . . . 12483 192 1 In in IN 12483 192 2 the the DT 12483 192 3 ( ( -LRB- 12483 192 4 BOW BOW NNP 12483 192 5 ) ) -RRB- 12483 192 6 model model NN 12483 192 7 , , , 12483 192 8 documents document NNS 12483 192 9 in in IN 12483 192 10 TRFN TRFN NNP 12483 192 11 are be VBP 12483 192 12 represented represent VBN 12483 192 13 as as IN 12483 192 14 a a DT 12483 192 15 collection collection NN 12483 192 16 of of IN 12483 192 17 words word NNS 12483 192 18 , , , 12483 192 19 ignoring ignore VBG 12483 192 20 grammar grammar NN 12483 192 21 and and CC 12483 192 22 even even RB 12483 192 23 word word NN 12483 192 24 order order NN 12483 192 25 , , , 12483 192 26 but but CC 12483 192 27 preserving preserve VBG 12483 192 28 multiplicity multiplicity NN 12483 192 29 . . . 12483 193 1 In in IN 12483 193 2 a a DT 12483 193 3 classic classic JJ 12483 193 4 BOW BOW NNP 12483 193 5 approach approach NN 12483 193 6 , , , 12483 193 7 each each DT 12483 193 8 document document NN 12483 193 9 can can MD 12483 193 10 be be VB 12483 193 11 represented represent VBN 12483 193 12 as as IN 12483 193 13 a a DT 12483 193 14 fixed fix VBN 12483 193 15 - - HYPH 12483 193 16 length length NN 12483 193 17 vector vector NN 12483 193 18 with with IN 12483 193 19 length length NN 12483 193 20 equal equal JJ 12483 193 21 to to IN 12483 193 22 the the DT 12483 193 23 vocabulary vocabulary NN 12483 193 24 size size NN 12483 193 25 . . . 12483 194 1 This this DT 12483 194 2 means mean VBZ 12483 194 3 each each DT 12483 194 4 dimension dimension NN 12483 194 5 of of IN 12483 194 6 this this DT 12483 194 7 vector vector NN 12483 194 8 corresponds correspond VBZ 12483 194 9 to to IN 12483 194 10 the the DT 12483 194 11 occurrence occurrence NN 12483 194 12 of of IN 12483 194 13 a a DT 12483 194 14 word word NN 12483 194 15 in in IN 12483 194 16 a a DT 12483 194 17 news news NN 12483 194 18 item item NN 12483 194 19 . . . 12483 195 1 We -PRON- PRP 12483 195 2 customized customize VBD 12483 195 3 the the DT 12483 195 4 generic generic JJ 12483 195 5 approach approach NN 12483 195 6 by by IN 12483 195 7 reducing reduce VBG 12483 195 8 variable variable JJ 12483 195 9 - - HYPH 12483 195 10 length length NN 12483 195 11 documents document NNS 12483 195 12 to to IN 12483 195 13 fixed fix VBN 12483 195 14 - - HYPH 12483 195 15 length length NN 12483 195 16 vectors vector NNS 12483 195 17 to to TO 12483 195 18 be be VB 12483 195 19 able able JJ 12483 195 20 to to TO 12483 195 21 use use VB 12483 195 22 with with IN 12483 195 23 varying varying NN 12483 195 24 lengths length NNS 12483 195 25 with with IN 12483 195 26 many many JJ 12483 195 27 machine machine NN 12483 195 28 learning learning NN 12483 195 29 models model NNS 12483 195 30 . . . 12483 196 1 INFORMATION INFORMATION NNP 12483 196 2 TECHNOLOGY technology NN 12483 196 3 AND and CC 12483 196 4 LIBRARIES library NNS 12483 196 5 DECEMBER DECEMBER NNP 12483 196 6 2020 2020 CD 12483 196 7 AUTOMATED AUTOMATED NNP 12483 196 8 FAKE FAKE NNP 12483 196 9 NEWS NEWS NNP 12483 196 10 DETECTION detection VB 12483 196 11 IN in IN 12483 196 12 THE the DT 12483 196 13 AGE AGE NNP 12483 196 14 OF of IN 12483 196 15 DIGITAL DIGITAL NNP 12483 196 16 LIBRARIES library NNS 12483 196 17 | | NNP 12483 196 18 MERTOĞLU MERTOĞLU NNS 12483 196 19 AND and CC 12483 196 20 GENÇ genç VBP 12483 196 21 13 13 CD 12483 196 22 Figure figure NN 12483 196 23 5 5 CD 12483 196 24 . . . 12483 197 1 An an DT 12483 197 2 overview overview NN 12483 197 3 of of IN 12483 197 4 BOW BOW NNP 12483 197 5 ( ( -LRB- 12483 197 6 Bag Bag NNP 12483 197 7 of of IN 12483 197 8 Word Word NNP 12483 197 9 ) ) -RRB- 12483 197 10 Approach Approach NNP 12483 197 11 . . . 12483 198 1 Because because IN 12483 198 2 we -PRON- PRP 12483 198 3 ignore ignore VBP 12483 198 4 the the DT 12483 198 5 word word NN 12483 198 6 order order NN 12483 198 7 , , , 12483 198 8 we -PRON- PRP 12483 198 9 reduced reduce VBD 12483 198 10 fixed fixed JJ 12483 198 11 length length NN 12483 198 12 of of IN 12483 198 13 counts count NNS 12483 198 14 as as IN 12483 198 15 histograms histogram NNS 12483 198 16 as as IN 12483 198 17 seen see VBN 12483 198 18 in in IN 12483 198 19 figure figure NN 12483 198 20 5 5 CD 12483 198 21 . . . 12483 199 1 Assuming assume VBG 12483 199 2 N N NNP 12483 199 3 is be VBZ 12483 199 4 the the DT 12483 199 5 number number NN 12483 199 6 of of IN 12483 199 7 news news NN 12483 199 8 documents document NNS 12483 199 9 and and CC 12483 199 10 W w NN 12483 199 11 is be VBZ 12483 199 12 the the DT 12483 199 13 number number NN 12483 199 14 of of IN 12483 199 15 possible possible JJ 12483 199 16 words word NNS 12483 199 17 in in IN 12483 199 18 the the DT 12483 199 19 corpus corpus NN 12483 199 20 , , , 12483 199 21 it -PRON- PRP 12483 199 22 should should MD 12483 199 23 be be VB 12483 199 24 noted note VBN 12483 199 25 that that IN 12483 199 26 in in IN 12483 199 27 N*W N*W NNP 12483 199 28 count count NNP 12483 199 29 matrix matrix NNP 12483 199 30 , , , 12483 199 31 N N NNP 12483 199 32 is be VBZ 12483 199 33 generally generally RB 12483 199 34 large large JJ 12483 199 35 but but CC 12483 199 36 infrequent infrequent JJ 12483 199 37 , , , 12483 199 38 because because IN 12483 199 39 we -PRON- PRP 12483 199 40 have have VBP 12483 199 41 many many JJ 12483 199 42 news news NN 12483 199 43 documents document NNS 12483 199 44 , , , 12483 199 45 but but CC 12483 199 46 most most JJS 12483 199 47 words word NNS 12483 199 48 do do VBP 12483 199 49 not not RB 12483 199 50 occur occur VB 12483 199 51 in in IN 12483 199 52 any any DT 12483 199 53 given give VBN 12483 199 54 document document NN 12483 199 55 causing cause VBG 12483 199 56 rareness rareness NN 12483 199 57 of of IN 12483 199 58 a a DT 12483 199 59 term term NN 12483 199 60 / / SYM 12483 199 61 word word NN 12483 199 62 which which WDT 12483 199 63 is be VBZ 12483 199 64 a a DT 12483 199 65 drawback drawback NN 12483 199 66 for for IN 12483 199 67 the the DT 12483 199 68 approach approach NN 12483 199 69 . . . 12483 200 1 Therefore therefore RB 12483 200 2 , , , 12483 200 3 we -PRON- PRP 12483 200 4 modified modify VBD 12483 200 5 the the DT 12483 200 6 model model NN 12483 200 7 to to TO 12483 200 8 compensate compensate VB 12483 200 9 the the DT 12483 200 10 rarity rarity NN 12483 200 11 problem problem NN 12483 200 12 by by IN 12483 200 13 weighting weight VBG 12483 200 14 the the DT 12483 200 15 terms term NNS 12483 200 16 using use VBG 12483 200 17 TF TF NNP 12483 200 18 - - HYPH 12483 200 19 IDF IDF NNP 12483 200 20 measure measure NN 12483 200 21 which which WDT 12483 200 22 evaluates evaluate VBZ 12483 200 23 how how WRB 12483 200 24 important important JJ 12483 200 25 a a DT 12483 200 26 word word NN 12483 200 27 is be VBZ 12483 200 28 to to IN 12483 200 29 a a DT 12483 200 30 document document NN 12483 200 31 in in IN 12483 200 32 a a DT 12483 200 33 collection collection NN 12483 200 34 . . . 12483 201 1 The the DT 12483 201 2 other other JJ 12483 201 3 technique technique NN 12483 201 4 we -PRON- PRP 12483 201 5 used use VBD 12483 201 6 , , , 12483 201 7 N n NN 12483 201 8 - - HYPH 12483 201 9 gram gram NN 12483 201 10 model model NN 12483 201 11 is be VBZ 12483 201 12 the the DT 12483 201 13 generic generic JJ 12483 201 14 term term NN 12483 201 15 for for IN 12483 201 16 a a DT 12483 201 17 string string NN 12483 201 18 of of IN 12483 201 19 words word NNS 12483 201 20 in in IN 12483 201 21 computational computational JJ 12483 201 22 linguistics linguistic NNS 12483 201 23 , , , 12483 201 24 and and CC 12483 201 25 it -PRON- PRP 12483 201 26 is be VBZ 12483 201 27 extensively extensively RB 12483 201 28 used use VBN 12483 201 29 in in IN 12483 201 30 text text NN 12483 201 31 mining mining NN 12483 201 32 and and CC 12483 201 33 NLP NLP NNP 12483 201 34 tasks task NNS 12483 201 35 . . . 12483 202 1 The the DT 12483 202 2 prefixes prefix NNS 12483 202 3 that that WDT 12483 202 4 replace replace VBP 12483 202 5 the the DT 12483 202 6 n n JJ 12483 202 7 - - HYPH 12483 202 8 part part NN 12483 202 9 indicate indicate VBP 12483 202 10 the the DT 12483 202 11 number number NN 12483 202 12 of of IN 12483 202 13 consecutive consecutive JJ 12483 202 14 words word NNS 12483 202 15 in in IN 12483 202 16 the the DT 12483 202 17 string string NN 12483 202 18 . . . 12483 203 1 So so RB 12483 203 2 , , , 12483 203 3 a a DT 12483 203 4 unigram unigram NN 12483 203 5 is be VBZ 12483 203 6 referred refer VBN 12483 203 7 to to IN 12483 203 8 one one CD 12483 203 9 word word NN 12483 203 10 , , , 12483 203 11 a a DT 12483 203 12 bigram bigram NN 12483 203 13 is be VBZ 12483 203 14 two two CD 12483 203 15 words word NNS 12483 203 16 , , , 12483 203 17 and and CC 12483 203 18 an an DT 12483 203 19 n n NN 12483 203 20 - - HYPH 12483 203 21 gram gram NN 12483 203 22 is be VBZ 12483 203 23 n n CD 12483 203 24 words word NNS 12483 203 25 . . . 12483 204 1 EXPERIMENTAL experimental NN 12483 204 2 RESULTS result NNS 12483 204 3 AND and CC 12483 204 4 DISCUSSION discussion NN 12483 204 5 In in IN 12483 204 6 this this DT 12483 204 7 section section NN 12483 204 8 , , , 12483 204 9 the the DT 12483 204 10 experimental experimental JJ 12483 204 11 process process NN 12483 204 12 and and CC 12483 204 13 the the DT 12483 204 14 results result NNS 12483 204 15 are be VBP 12483 204 16 presented present VBN 12483 204 17 . . . 12483 205 1 All all DT 12483 205 2 experiments experiment NNS 12483 205 3 are be VBP 12483 205 4 performed perform VBN 12483 205 5 using use VBG 12483 205 6 the the DT 12483 205 7 Scikit Scikit NNPS 12483 205 8 - - HYPH 12483 205 9 learn learn VB 12483 205 10 library library NN 12483 205 11 . . . 12483 206 1 To to TO 12483 206 2 evaluate evaluate VB 12483 206 3 the the DT 12483 206 4 performance performance NN 12483 206 5 of of IN 12483 206 6 the the DT 12483 206 7 model model NN 12483 206 8 and and CC 12483 206 9 proposed propose VBN 12483 206 10 features feature NNS 12483 206 11 we -PRON- PRP 12483 206 12 employed employ VBD 12483 206 13 the the DT 12483 206 14 precision precision NN 12483 206 15 , , , 12483 206 16 recall recall NN 12483 206 17 , , , 12483 206 18 F1 F1 NNP 12483 206 19 score score NN 12483 206 20 ( ( -LRB- 12483 206 21 the the DT 12483 206 22 harmonic harmonic JJ 12483 206 23 mean mean NN 12483 206 24 of of IN 12483 206 25 the the DT 12483 206 26 precision precision NN 12483 206 27 and and CC 12483 206 28 recall recall NN 12483 206 29 ) ) -RRB- 12483 206 30 , , , 12483 206 31 and and CC 12483 206 32 accuracy accuracy NN 12483 206 33 metrics metric NNS 12483 206 34 . . . 12483 207 1 We -PRON- PRP 12483 207 2 did do VBD 12483 207 3 many many JJ 12483 207 4 experiments experiment NNS 12483 207 5 using use VBG 12483 207 6 different different JJ 12483 207 7 combinations combination NNS 12483 207 8 of of IN 12483 207 9 features feature NNS 12483 207 10 . . . 12483 208 1 INFORMATION INFORMATION NNP 12483 208 2 TECHNOLOGY technology NN 12483 208 3 AND and CC 12483 208 4 LIBRARIES library NNS 12483 208 5 DECEMBER DECEMBER NNP 12483 208 6 2020 2020 CD 12483 208 7 AUTOMATED AUTOMATED NNP 12483 208 8 FAKE FAKE NNP 12483 208 9 NEWS NEWS NNP 12483 208 10 DETECTION detection VB 12483 208 11 IN in IN 12483 208 12 THE the DT 12483 208 13 AGE AGE NNP 12483 208 14 OF of IN 12483 208 15 DIGITAL DIGITAL NNP 12483 208 16 LIBRARIES library NNS 12483 208 17 | | NNP 12483 208 18 MERTOĞLU MERTOĞLU NNS 12483 208 19 AND and CC 12483 208 20 GENÇ genç NN 12483 208 21 14 14 CD 12483 208 22 Several several JJ 12483 208 23 classification classification NN 12483 208 24 models model NNS 12483 208 25 have have VBP 12483 208 26 been be VBN 12483 208 27 trained train VBN 12483 208 28 . . . 12483 209 1 These these DT 12483 209 2 are be VBP 12483 209 3 as as IN 12483 209 4 follows follow VBZ 12483 209 5 : : : 12483 209 6 K K NNP 12483 209 7 - - HYPH 12483 209 8 Nearest Nearest NNP 12483 209 9 Neighbor Neighbor NNP 12483 209 10 , , , 12483 209 11 Decision Decision NNP 12483 209 12 Trees Trees NNP 12483 209 13 , , , 12483 209 14 Gaussian Gaussian NNP 12483 209 15 Naive Naive NNP 12483 209 16 Bayes Bayes NNP 12483 209 17 , , , 12483 209 18 Random Random NNP 12483 209 19 Forest Forest NNP 12483 209 20 , , , 12483 209 21 Support Support NNP 12483 209 22 Vector Vector NNP 12483 209 23 Machine Machine NNP 12483 209 24 , , , 12483 209 25 ExtraTrees ExtraTrees NNP 12483 209 26 Classifier Classifier NNP 12483 209 27 , , , 12483 209 28 and and CC 12483 209 29 Logistic Logistic NNP 12483 209 30 Regression Regression NNP 12483 209 31 . . . 12483 210 1 To to TO 12483 210 2 be be VB 12483 210 3 effective effective JJ 12483 210 4 , , , 12483 210 5 a a DT 12483 210 6 classifier classifier NN 12483 210 7 should should MD 12483 210 8 be be VB 12483 210 9 able able JJ 12483 210 10 to to TO 12483 210 11 correctly correctly RB 12483 210 12 classify classify VB 12483 210 13 previously previously RB 12483 210 14 unseen unseen JJ 12483 210 15 data datum NNS 12483 210 16 . . . 12483 211 1 To to IN 12483 211 2 this this DT 12483 211 3 end end NN 12483 211 4 , , , 12483 211 5 we -PRON- PRP 12483 211 6 tuned tune VBD 12483 211 7 the the DT 12483 211 8 parameter parameter NN 12483 211 9 values value NNS 12483 211 10 for for IN 12483 211 11 all all PDT 12483 211 12 the the DT 12483 211 13 classification classification NN 12483 211 14 models model NNS 12483 211 15 used use VBN 12483 211 16 . . . 12483 212 1 Then then RB 12483 212 2 , , , 12483 212 3 models model NNS 12483 212 4 were be VBD 12483 212 5 trained train VBN 12483 212 6 and and CC 12483 212 7 evaluated evaluate VBN 12483 212 8 on on IN 12483 212 9 TRFN TRFN NNP 12483 212 10 dataset dataset NN 12483 212 11 using use VBG 12483 212 12 10-fold 10-fold CD 12483 212 13 cross cross NN 12483 212 14 - - NN 12483 212 15 validation validation NN 12483 212 16 . . . 12483 213 1 In in IN 12483 213 2 table table NN 12483 213 3 3 3 CD 12483 213 4 , , , 12483 213 5 we -PRON- PRP 12483 213 6 present present VBP 12483 213 7 the the DT 12483 213 8 ultimate ultimate JJ 12483 213 9 best good JJS 12483 213 10 scores score NNS 12483 213 11 of of IN 12483 213 12 the the DT 12483 213 13 proposed propose VBN 12483 213 14 model model NN 12483 213 15 . . . 12483 214 1 The the DT 12483 214 2 results result NNS 12483 214 3 are be VBP 12483 214 4 highly highly RB 12483 214 5 motivating motivate VBG 12483 214 6 to to TO 12483 214 7 exemplify exemplify VB 12483 214 8 how how WRB 12483 214 9 useful useful JJ 12483 214 10 automated automated JJ 12483 214 11 detection detection NN 12483 214 12 systems system NNS 12483 214 13 can can MD 12483 214 14 be be VB 12483 214 15 as as IN 12483 214 16 a a DT 12483 214 17 key key JJ 12483 214 18 component component NN 12483 214 19 of of IN 12483 214 20 the the DT 12483 214 21 integrated integrate VBN 12483 214 22 solution solution NN 12483 214 23 framework framework NN 12483 214 24 in in IN 12483 214 25 figure figure NN 12483 214 26 2 2 CD 12483 214 27 . . . 12483 215 1 We -PRON- PRP 12483 215 2 compared compare VBD 12483 215 3 the the DT 12483 215 4 algorithms algorithm NNS 12483 215 5 with with IN 12483 215 6 three three CD 12483 215 7 ultimate ultimate JJ 12483 215 8 feature feature NN 12483 215 9 sets set NNS 12483 215 10 for for IN 12483 215 11 having have VBG 12483 215 12 respectively respectively RB 12483 215 13 consistent consistent JJ 12483 215 14 results result NNS 12483 215 15 to to IN 12483 215 16 the the DT 12483 215 17 other other JJ 12483 215 18 feature feature NN 12483 215 19 set set VBN 12483 215 20 combinations combination NNS 12483 215 21 . . . 12483 216 1 Set1 set1 RB 12483 216 2 stands stand VBZ 12483 216 3 for for IN 12483 216 4 bigram+FOpt bigram+FOpt NNP 12483 216 5 ( ( -LRB- 12483 216 6 Optimized Optimized NNP 12483 216 7 Features Features NNP 12483 216 8 ) ) -RRB- 12483 216 9 , , , 12483 216 10 Set2 Set2 NNP 12483 216 11 stands stand VBZ 12483 216 12 for for IN 12483 216 13 BOWModified+ BOWModified+ NNP 12483 216 14 FOpt FOpt NNP 12483 216 15 and and CC 12483 216 16 Set3 Set3 NNP 12483 216 17 stands stand VBZ 12483 216 18 for for IN 12483 216 19 unigram+bigram+FOpt unigram+bigram+FOpt NNS 12483 216 20 . . . 12483 217 1 The the DT 12483 217 2 results result NNS 12483 217 3 show show VBP 12483 217 4 that that IN 12483 217 5 there there EX 12483 217 6 is be VBZ 12483 217 7 a a DT 12483 217 8 relative relative JJ 12483 217 9 consistency consistency NN 12483 217 10 in in IN 12483 217 11 terms term NNS 12483 217 12 of of IN 12483 217 13 performance performance NN 12483 217 14 across across IN 12483 217 15 the the DT 12483 217 16 models model NNS 12483 217 17 . . . 12483 218 1 In in IN 12483 218 2 almost almost RB 12483 218 3 all all DT 12483 218 4 models model NNS 12483 218 5 , , , 12483 218 6 the the DT 12483 218 7 combination combination NN 12483 218 8 of of IN 12483 218 9 unigram+bigram unigram+bigram NNP 12483 218 10 and and CC 12483 218 11 optimized optimize VBN 12483 218 12 features feature NNS 12483 218 13 sets set NNS 12483 218 14 ( ( -LRB- 12483 218 15 FOpt FOpt NNP 12483 218 16 ) ) -RRB- 12483 218 17 gives give VBZ 12483 218 18 better well JJR 12483 218 19 results result NNS 12483 218 20 than than IN 12483 218 21 the the DT 12483 218 22 other other JJ 12483 218 23 combinations combination NNS 12483 218 24 . . . 12483 219 1 The the DT 12483 219 2 ExtraTree ExtraTree NNP 12483 219 3 Classifier Classifier NNP 12483 219 4 model model NN 12483 219 5 is be VBZ 12483 219 6 chosen choose VBN 12483 219 7 as as IN 12483 219 8 the the DT 12483 219 9 best good JJS 12483 219 10 due due JJ 12483 219 11 to to IN 12483 219 12 its -PRON- PRP$ 12483 219 13 higher high JJR 12483 219 14 performance performance NN 12483 219 15 . . . 12483 220 1 This this DT 12483 220 2 model model NN 12483 220 3 is be VBZ 12483 220 4 also also RB 12483 220 5 known know VBN 12483 220 6 as as IN 12483 220 7 Extremely extremely RB 12483 220 8 Randomized randomized JJ 12483 220 9 Trees tree NNS 12483 220 10 Classifier Classifier NNP 12483 220 11 which which WDT 12483 220 12 is be VBZ 12483 220 13 a a DT 12483 220 14 type type NN 12483 220 15 of of IN 12483 220 16 ensemble ensemble JJ 12483 220 17 learning learning NN 12483 220 18 technique technique NN 12483 220 19 aggregating aggregate VBG 12483 220 20 the the DT 12483 220 21 results result NNS 12483 220 22 of of IN 12483 220 23 multiple multiple JJ 12483 220 24 decision decision NN 12483 220 25 trees tree NNS 12483 220 26 collected collect VBN 12483 220 27 in in IN 12483 220 28 a a DT 12483 220 29 “ " `` 12483 220 30 forest forest NN 12483 220 31 ” " '' 12483 220 32 to to TO 12483 220 33 output output VB 12483 220 34 its -PRON- PRP$ 12483 220 35 classification classification NN 12483 220 36 result result NN 12483 220 37 . . . 12483 221 1 It -PRON- PRP 12483 221 2 is be VBZ 12483 221 3 very very RB 12483 221 4 similar similar JJ 12483 221 5 to to IN 12483 221 6 Random Random NNP 12483 221 7 Forest Forest NNP 12483 221 8 Classifier Classifier NNP 12483 221 9 and and CC 12483 221 10 only only RB 12483 221 11 differs differ VBZ 12483 221 12 in in IN 12483 221 13 the the DT 12483 221 14 manner manner NN 12483 221 15 of of IN 12483 221 16 construction construction NN 12483 221 17 of of IN 12483 221 18 the the DT 12483 221 19 decision decision NN 12483 221 20 trees tree NNS 12483 221 21 . . . 12483 222 1 So so CC 12483 222 2 , , , 12483 222 3 we -PRON- PRP 12483 222 4 can can MD 12483 222 5 also also RB 12483 222 6 see see VB 12483 222 7 closer close JJR 12483 222 8 results result NNS 12483 222 9 between between IN 12483 222 10 these these DT 12483 222 11 two two CD 12483 222 12 classifiers classifier NNS 12483 222 13 . . . 12483 223 1 Table table NN 12483 223 2 3 3 CD 12483 223 3 . . . 12483 224 1 Results result NNS 12483 224 2 . . . 12483 225 1 Evaluation evaluation NN 12483 225 2 results result NNS 12483 225 3 of of IN 12483 225 4 all all DT 12483 225 5 combinations combination NNS 12483 225 6 of of IN 12483 225 7 features feature NNS 12483 225 8 and and CC 12483 225 9 classification classification NN 12483 225 10 models model NNS 12483 225 11 . . . 12483 226 1 Model model NN 12483 226 2 Feature feature NN 12483 226 3 Sets set NNS 12483 226 4 Precision%(0,1 precision%(0,1 IN 12483 226 5 ) ) -RRB- 12483 226 6 Recall%(0,1 Recall%(0,1 NFP 12483 226 7 ) ) -RRB- 12483 226 8 Accurac Accurac NNP 12483 226 9 y y NNP 12483 226 10 F1Scor f1scor NN 12483 226 11 e e NNP 12483 226 12 Set1 Set1 NNP 12483 226 13 93.32 93.32 CD 12483 226 14 93.96 93.96 CD 12483 226 15 93.92 93.92 CD 12483 226 16 93.3 93.3 CD 12483 226 17 6 6 CD 12483 226 18 93.64 93.64 CD 12483 226 19 93.62 93.62 CD 12483 226 20 Gaussian Gaussian NNP 12483 226 21 Naive Naive NNP 12483 226 22 Bayes Bayes NNP 12483 226 23 Set2 set2 NN 12483 226 24 93.37 93.37 CD 12483 226 25 94.02 94.02 CD 12483 226 26 93.98 93.98 CD 12483 226 27 93.4 93.4 CD 12483 226 28 2 2 CD 12483 226 29 93.70 93.70 CD 12483 226 30 93.68 93.68 CD 12483 226 31 Set3 set3 CD 12483 226 32 93.95 93.95 CD 12483 226 33 94.21 94.21 CD 12483 226 34 94.19 94.19 CD 12483 226 35 93.9 93.9 CD 12483 226 36 7 7 CD 12483 226 37 94.08 94.08 CD 12483 226 38 94.07 94.07 CD 12483 226 39 Set1 set1 CD 12483 226 40 93.70 93.70 CD 12483 226 41 93.50 93.50 CD 12483 226 42 93.52 93.52 CD 12483 226 43 93.6 93.6 CD 12483 226 44 9 9 CD 12483 226 45 93.60 93.60 CD 12483 226 46 93.61 93.61 CD 12483 226 47 K K NNP 12483 226 48 - - HYPH 12483 226 49 Nearest Nearest NNP 12483 226 50 Neighbour Neighbour NNP 12483 226 51 Set2 Set2 NNP 12483 226 52 93.66 93.66 CD 12483 226 53 94.05 94.05 CD 12483 226 54 94.03 94.03 CD 12483 226 55 93.6 93.6 CD 12483 226 56 8 8 CD 12483 226 57 93.85 93.85 CD 12483 226 58 93.84 93.84 CD 12483 226 59 Set3 set3 CD 12483 226 60 94.42 94.42 CD 12483 226 61 94.21 94.21 CD 12483 226 62 94.22 94.22 CD 12483 226 63 94.4 94.4 CD 12483 226 64 1 1 CD 12483 226 65 94.31 94.31 CD 12483 226 66 94.32 94.32 CD 12483 226 67 Set1 Set1 NNP 12483 226 68 94.15 94.15 CD 12483 226 69 94.92 94.92 CD 12483 226 70 94.88 94.88 CD 12483 226 71 94.1 94.1 CD 12483 226 72 9 9 CD 12483 226 73 94.53 94.53 CD 12483 226 74 94.51 94.51 CD 12483 226 75 ExtraTrees ExtraTrees NNP 12483 226 76 Classifier Classifier NNP 12483 226 77 Set2 Set2 NNP 12483 226 78 94.09 94.09 CD 12483 226 79 94.94 94.94 CD 12483 226 80 94.90 94.90 CD 12483 226 81 94.1 94.1 CD 12483 226 82 4 4 CD 12483 226 83 94.51 94.51 CD 12483 226 84 94.49 94.49 CD 12483 226 85 Set3 Set3 NNP 12483 226 86 97.90 97.90 CD 12483 226 87 95.72 95.72 CD 12483 226 88 95.81 95.81 CD 12483 226 89 97.8 97.8 CD 12483 226 90 6 6 CD 12483 226 91 96.81 96.81 CD 12483 226 92 96.85 96.85 CD 12483 226 93 Set1 set1 CD 12483 226 94 89.61 89.61 CD 12483 226 95 88.92 88.92 CD 12483 226 96 88.99 88.99 CD 12483 226 97 89.5 89.5 CD 12483 226 98 4 4 CD 12483 226 99 89.26 89.26 CD 12483 226 100 89.30 89.30 CD 12483 226 101 Support Support NNP 12483 226 102 Vector Vector NNP 12483 226 103 Machine Machine NNP 12483 226 104 Set2 Set2 NNP 12483 226 105 89.70 89.70 CD 12483 226 106 88.96 88.96 CD 12483 226 107 89.04 89.04 CD 12483 226 108 89.6 89.6 CD 12483 226 109 2 2 CD 12483 226 110 89.33 89.33 CD 12483 226 111 89.37 89.37 CD 12483 226 112 Set3 set3 CD 12483 226 113 90.85 90.85 CD 12483 226 114 91.26 91.26 CD 12483 226 115 91.22 91.22 CD 12483 226 116 90.8 90.8 CD 12483 226 117 9 9 CD 12483 226 118 91.05 91.05 CD 12483 226 119 91.03 91.03 CD 12483 226 120 Set1 set1 CD 12483 226 121 91.56 91.56 CD 12483 226 122 92.28 92.28 CD 12483 226 123 92.23 92.23 CD 12483 226 124 91.6 91.6 CD 12483 226 125 2 2 CD 12483 226 126 91.92 91.92 CD 12483 226 127 91.89 91.89 CD 12483 226 128 Logistic Logistic NNP 12483 226 129 Regression Regression NNP 12483 226 130 Set2 Set2 NNP 12483 226 131 91.50 91.50 CD 12483 226 132 92.28 92.28 CD 12483 226 133 92.22 92.22 CD 12483 226 134 91.5 91.5 CD 12483 226 135 6 6 CD 12483 226 136 91.89 91.89 CD 12483 226 137 91.86 91.86 CD 12483 226 138 Set3 Set3 NNP 12483 226 139 92.25 92.25 CD 12483 226 140 92.90 92.90 CD 12483 226 141 92.86 92.86 CD 12483 226 142 92.3 92.3 CD 12483 226 143 0 0 CD 12483 226 144 92.57 92.57 CD 12483 226 145 92.55 92.55 CD 12483 226 146 Set1 set1 CD 12483 226 147 93.71 93.71 CD 12483 226 148 94.44 94.44 CD 12483 226 149 94.40 94.40 CD 12483 226 150 93.7 93.7 CD 12483 226 151 5 5 CD 12483 226 152 94.07 94.07 CD 12483 226 153 94.05 94.05 CD 12483 226 154 Random Random NNP 12483 226 155 Forest Forest NNP 12483 226 156 Set2 Set2 NNP 12483 226 157 93.87 93.87 CD 12483 226 158 95.00 95.00 CD 12483 226 159 94.94 94.94 CD 12483 226 160 93.9 93.9 CD 12483 226 161 4 4 CD 12483 226 162 94.44 94.44 CD 12483 226 163 94.41 94.41 CD 12483 226 164 Set3 set3 CD 12483 226 165 94.77 94.77 CD 12483 226 166 95.14 95.14 CD 12483 226 167 95.12 95.12 CD 12483 226 168 94.7 94.7 CD 12483 226 169 9 9 CD 12483 226 170 94.96 94.96 CD 12483 226 171 94.95 94.95 CD 12483 226 172 Set1 set1 CD 12483 226 173 93.95 93.95 CD 12483 226 174 94.59 94.59 CD 12483 226 175 94.56 94.56 CD 12483 226 176 93.9 93.9 CD 12483 226 177 9 9 CD 12483 226 178 94.27 94.27 CD 12483 226 179 94.25 94.25 CD 12483 226 180 Decision Decision NNP 12483 226 181 Trees Trees NNP 12483 226 182 Set2 Set2 NNP 12483 226 183 94.05 94.05 CD 12483 226 184 95.08 95.08 CD 12483 226 185 95.03 95.03 CD 12483 226 186 94.1 94.1 CD 12483 226 187 1 1 CD 12483 226 188 94.57 94.57 CD 12483 226 189 94.54 94.54 CD 12483 226 190 Set3 Set3 NNP 12483 226 191 94.94 94.94 CD 12483 226 192 95.24 95.24 CD 12483 226 193 95.23 95.23 CD 12483 226 194 94.9 94.9 CD 12483 226 195 5 5 CD 12483 226 196 95.09 95.09 CD 12483 226 197 95.08 95.08 CD 12483 226 198 INFORMATION INFORMATION NNP 12483 226 199 TECHNOLOGY technology NN 12483 226 200 AND and CC 12483 226 201 LIBRARIES library NNS 12483 226 202 DECEMBER DECEMBER NNP 12483 226 203 2020 2020 CD 12483 226 204 AUTOMATED AUTOMATED NNP 12483 226 205 FAKE FAKE NNP 12483 226 206 NEWS NEWS NNP 12483 226 207 DETECTION detection VB 12483 226 208 IN in IN 12483 226 209 THE the DT 12483 226 210 AGE AGE NNP 12483 226 211 OF of IN 12483 226 212 DIGITAL DIGITAL NNP 12483 226 213 LIBRARIES library NNS 12483 226 214 | | NNP 12483 226 215 MERTOĞLU MERTOĞLU NNS 12483 226 216 AND and CC 12483 226 217 GENÇ genç VBP 12483 226 218 15 15 CD 12483 226 219 Every every DT 12483 226 220 ADS ads NN 12483 226 221 in in IN 12483 226 222 GLIS_1.0 glis_1.0 CD 12483 226 223 framework framework NN 12483 226 224 may may MD 12483 226 225 use use VB 12483 226 226 its -PRON- PRP$ 12483 226 227 own own JJ 12483 226 228 way way NN 12483 226 229 to to TO 12483 226 230 detect detect VB 12483 226 231 fake fake JJ 12483 226 232 news news NN 12483 226 233 . . . 12483 227 1 The the DT 12483 227 2 open open JJ 12483 227 3 source source NN 12483 227 4 ADS ADS NNP 12483 227 5 may may MD 12483 227 6 improve improve VB 12483 227 7 with with IN 12483 227 8 feedbacks feedback NNS 12483 227 9 . . . 12483 228 1 Hybrid hybrid JJ 12483 228 2 models model NNS 12483 228 3 and and CC 12483 228 4 other other JJ 12483 228 5 techniques technique NNS 12483 228 6 such such JJ 12483 228 7 as as IN 12483 228 8 neural neural JJ 12483 228 9 networks network NNS 12483 228 10 with with IN 12483 228 11 deep deep JJ 12483 228 12 learning learning NN 12483 228 13 methodology methodology NN 12483 228 14 can can MD 12483 228 15 also also RB 12483 228 16 be be VB 12483 228 17 used use VBN 12483 228 18 according accord VBG 12483 228 19 to to IN 12483 228 20 the the DT 12483 228 21 data data NN 12483 228 22 , , , 12483 228 23 language language NN 12483 228 24 of of IN 12483 228 25 news news NN 12483 228 26 and and CC 12483 228 27 the the DT 12483 228 28 news news NN 12483 228 29 features feature VBZ 12483 228 30 related relate VBN 12483 228 31 with with IN 12483 228 32 both both CC 12483 228 33 social social JJ 12483 228 34 context context NN 12483 228 35 and and CC 12483 228 36 news news NN 12483 228 37 content content NN 12483 228 38 . . . 12483 229 1 CONCLUSION conclusion NN 12483 229 2 AND and CC 12483 229 3 FUTURE future NN 12483 229 4 WORK work VBP 12483 229 5 In in IN 12483 229 6 this this DT 12483 229 7 study study NN 12483 229 8 we -PRON- PRP 12483 229 9 presented present VBD 12483 229 10 a a DT 12483 229 11 novel novel JJ 12483 229 12 framework framework NN 12483 229 13 which which WDT 12483 229 14 offers offer VBZ 12483 229 15 a a DT 12483 229 16 practical practical JJ 12483 229 17 architecture architecture NN 12483 229 18 of of IN 12483 229 19 an an DT 12483 229 20 integrated integrated JJ 12483 229 21 system system NN 12483 229 22 for for IN 12483 229 23 identifying identify VBG 12483 229 24 fake fake JJ 12483 229 25 news news NN 12483 229 26 . . . 12483 230 1 We -PRON- PRP 12483 230 2 have have VBP 12483 230 3 tried try VBN 12483 230 4 to to TO 12483 230 5 illustrate illustrate VB 12483 230 6 how how WRB 12483 230 7 digital digital JJ 12483 230 8 libraries library NNS 12483 230 9 can can MD 12483 230 10 be be VB 12483 230 11 a a DT 12483 230 12 service service NN 12483 230 13 authority authority NN 12483 230 14 to to TO 12483 230 15 promote promote VB 12483 230 16 media medium NNS 12483 230 17 literacy literacy NN 12483 230 18 and and CC 12483 230 19 fight fight VB 12483 230 20 against against IN 12483 230 21 fake fake JJ 12483 230 22 news news NN 12483 230 23 . . . 12483 231 1 Because because IN 12483 231 2 librarians librarian NNS 12483 231 3 are be VBP 12483 231 4 trained train VBN 12483 231 5 to to TO 12483 231 6 critically critically RB 12483 231 7 analyze analyze VB 12483 231 8 information information NN 12483 231 9 sources source NNS 12483 231 10 , , , 12483 231 11 their -PRON- PRP$ 12483 231 12 contributions contribution NNS 12483 231 13 to to IN 12483 231 14 our -PRON- PRP$ 12483 231 15 proposed propose VBN 12483 231 16 model model NN 12483 231 17 are be VBP 12483 231 18 critical critical JJ 12483 231 19 . . . 12483 232 1 Accordingly accordingly RB 12483 232 2 , , , 12483 232 3 we -PRON- PRP 12483 232 4 see see VBP 12483 232 5 this this DT 12483 232 6 work work NN 12483 232 7 as as IN 12483 232 8 an an DT 12483 232 9 encouraging encouraging JJ 12483 232 10 effort effort NN 12483 232 11 for for IN 12483 232 12 the the DT 12483 232 13 next next JJ 12483 232 14 collaborative collaborative JJ 12483 232 15 studies study NNS 12483 232 16 among among IN 12483 232 17 the the DT 12483 232 18 communities community NNS 12483 232 19 of of IN 12483 232 20 LIS LIS NNP 12483 232 21 and and CC 12483 232 22 CS CS NNP 12483 232 23 ( ( -LRB- 12483 232 24 computer computer NN 12483 232 25 science science NN 12483 232 26 ) ) -RRB- 12483 232 27 . . . 12483 233 1 We -PRON- PRP 12483 233 2 think think VBP 12483 233 3 that that IN 12483 233 4 there there EX 12483 233 5 is be VBZ 12483 233 6 an an DT 12483 233 7 immediate immediate JJ 12483 233 8 need need NN 12483 233 9 for for IN 12483 233 10 LIS LIS NNP 12483 233 11 professionals professional NNS 12483 233 12 to to TO 12483 233 13 participate participate VB 12483 233 14 and and CC 12483 233 15 contribute contribute VB 12483 233 16 to to IN 12483 233 17 automated automate VBN 12483 233 18 solutions solution NNS 12483 233 19 that that WDT 12483 233 20 can can MD 12483 233 21 help help VB 12483 233 22 detecting detect VBG 12483 233 23 inaccurate inaccurate JJ 12483 233 24 and and CC 12483 233 25 unverified unverified JJ 12483 233 26 information information NN 12483 233 27 . . . 12483 234 1 In in IN 12483 234 2 the the DT 12483 234 3 same same JJ 12483 234 4 manner manner NN 12483 234 5 , , , 12483 234 6 we -PRON- PRP 12483 234 7 believe believe VBP 12483 234 8 the the DT 12483 234 9 collaboration collaboration NN 12483 234 10 of of IN 12483 234 11 LIS LIS NNP 12483 234 12 professionals professional NNS 12483 234 13 , , , 12483 234 14 computer computer NN 12483 234 15 scientists scientist NNS 12483 234 16 , , , 12483 234 17 fact fact NN 12483 234 18 - - HYPH 12483 234 19 checking check VBG 12483 234 20 organizations organization NNS 12483 234 21 , , , 12483 234 22 and and CC 12483 234 23 pioneering pioneer VBG 12483 234 24 technology technology NN 12483 234 25 platforms platform NNS 12483 234 26 is be VBZ 12483 234 27 the the DT 12483 234 28 key key NN 12483 234 29 to to TO 12483 234 30 provide provide VB 12483 234 31 qualified qualified JJ 12483 234 32 news news NN 12483 234 33 within within IN 12483 234 34 a a DT 12483 234 35 real real JJ 12483 234 36 - - HYPH 12483 234 37 time time NN 12483 234 38 framework framework NN 12483 234 39 to to TO 12483 234 40 promote promote VB 12483 234 41 information information NN 12483 234 42 literacy literacy NN 12483 234 43 . . . 12483 235 1 Moreover moreover RB 12483 235 2 , , , 12483 235 3 we -PRON- PRP 12483 235 4 put put VBD 12483 235 5 the the DT 12483 235 6 reader reader NN 12483 235 7 at at IN 12483 235 8 the the DT 12483 235 9 core core NN 12483 235 10 of of IN 12483 235 11 the the DT 12483 235 12 framework framework NN 12483 235 13 as as IN 12483 235 14 the the DT 12483 235 15 feed feed NN 12483 235 16 reader reader NN 12483 235 17 position position NN 12483 235 18 while while IN 12483 235 19 consuming consume VBG 12483 235 20 news news NN 12483 235 21 . . . 12483 236 1 In in IN 12483 236 2 terms term NNS 12483 236 3 of of IN 12483 236 4 automated automate VBN 12483 236 5 detection detection NN 12483 236 6 systems system NNS 12483 236 7 , , , 12483 236 8 we -PRON- PRP 12483 236 9 proposed propose VBD 12483 236 10 a a DT 12483 236 11 fake fake JJ 12483 236 12 news news NN 12483 236 13 detection detection NN 12483 236 14 model model NN 12483 236 15 in in IN 12483 236 16 tegration tegration NN 12483 236 17 of of IN 12483 236 18 dictionary dictionary NNP 12483 236 19 - - HYPH 12483 236 20 based base VBN 12483 236 21 approach approach NN 12483 236 22 and and CC 12483 236 23 machine machine NN 12483 236 24 learning learning NN 12483 236 25 techniques technique NNS 12483 236 26 offering offer VBG 12483 236 27 optimized optimize VBN 12483 236 28 feature feature NN 12483 236 29 sets set NNS 12483 236 30 applicable applicable JJ 12483 236 31 to to IN 12483 236 32 agglutinative agglutinative JJ 12483 236 33 languages language NNS 12483 236 34 . . . 12483 237 1 We -PRON- PRP 12483 237 2 comparatively comparatively RB 12483 237 3 analyzed analyze VBD 12483 237 4 the the DT 12483 237 5 findings finding NNS 12483 237 6 with with IN 12483 237 7 several several JJ 12483 237 8 classification classification NN 12483 237 9 models model NNS 12483 237 10 . . . 12483 238 1 We -PRON- PRP 12483 238 2 demonstrated demonstrate VBD 12483 238 3 that that DT 12483 238 4 machine machine NN 12483 238 5 learning learn VBG 12483 238 6 algorithms algorithm NNS 12483 238 7 when when WRB 12483 238 8 used use VBN 12483 238 9 together together RB 12483 238 10 with with IN 12483 238 11 dictionary dictionary NNP 12483 238 12 - - HYPH 12483 238 13 based base VBN 12483 238 14 findings finding NNS 12483 238 15 yield yield VBP 12483 238 16 high high JJ 12483 238 17 scores score NNS 12483 238 18 both both DT 12483 238 19 for for IN 12483 238 20 precision precision NN 12483 238 21 and and CC 12483 238 22 recall recall NN 12483 238 23 . . . 12483 239 1 Consequently consequently RB 12483 239 2 , , , 12483 239 3 we -PRON- PRP 12483 239 4 believe believe VBP 12483 239 5 once once RB 12483 239 6 operational operational JJ 12483 239 7 in in IN 12483 239 8 the the DT 12483 239 9 field field NN 12483 239 10 , , , 12483 239 11 proposed propose VBN 12483 239 12 workflow workflow NN 12483 239 13 can can MD 12483 239 14 be be VB 12483 239 15 extended extend VBN 12483 239 16 in in IN 12483 239 17 the the DT 12483 239 18 future future NN 12483 239 19 to to TO 12483 239 20 support support VB 12483 239 21 other other JJ 12483 239 22 news news NN 12483 239 23 elements element NNS 12483 239 24 such such JJ 12483 239 25 as as IN 12483 239 26 photographs photograph NNS 12483 239 27 and and CC 12483 239 28 videos video NNS 12483 239 29 . . . 12483 240 1 With with IN 12483 240 2 the the DT 12483 240 3 help help NN 12483 240 4 of of IN 12483 240 5 Social Social NNP 12483 240 6 Network Network NNP 12483 240 7 Analysis Analysis NNP 12483 240 8 ( ( -LRB- 12483 240 9 SNA SNA NNP 12483 240 10 ) ) -RRB- 12483 240 11 it -PRON- PRP 12483 240 12 may may MD 12483 240 13 be be VB 12483 240 14 possible possible JJ 12483 240 15 to to TO 12483 240 16 stop stop VB 12483 240 17 or or CC 12483 240 18 slow slow VB 12483 240 19 down down RP 12483 240 20 the the DT 12483 240 21 spread spread NN 12483 240 22 of of IN 12483 240 23 fake fake JJ 12483 240 24 news news NN 12483 240 25 as as IN 12483 240 26 it -PRON- PRP 12483 240 27 emerges emerge VBZ 12483 240 28 . . . 12483 241 1 During during IN 12483 241 2 all all PDT 12483 241 3 the the DT 12483 241 4 experiments experiment NNS 12483 241 5 we -PRON- PRP 12483 241 6 did do VBD 12483 241 7 , , , 12483 241 8 this this DT 12483 241 9 work work NN 12483 241 10 also also RB 12483 241 11 highlighted highlight VBD 12483 241 12 several several JJ 12483 241 13 tasks task NNS 12483 241 14 as as IN 12483 241 15 future future JJ 12483 241 16 research research NN 12483 241 17 directions direction NNS 12483 241 18 such such JJ 12483 241 19 as as IN 12483 241 20 : : : 12483 241 21 • • VB 12483 241 22 The the DT 12483 241 23 studies study NNS 12483 241 24 can can MD 12483 241 25 be be VB 12483 241 26 deepened deepen VBN 12483 241 27 to to TO 12483 241 28 mathematically mathematically RB 12483 241 29 categorize categorize VB 12483 241 30 the the DT 12483 241 31 fake fake JJ 12483 241 32 news news NN 12483 241 33 types type NNS 12483 241 34 and and CC 12483 241 35 the the DT 12483 241 36 dissemination dissemination NN 12483 241 37 characteristics characteristic NNS 12483 241 38 of of IN 12483 241 39 each each DT 12483 241 40 type type NN 12483 241 41 can can MD 12483 241 42 be be VB 12483 241 43 analyzed analyze VBN 12483 241 44 . . . 12483 242 1 • • VB 12483 242 2 The the DT 12483 242 3 workflow workflow NN 12483 242 4 has have VBZ 12483 242 5 the the DT 12483 242 6 potential potential NN 12483 242 7 to to TO 12483 242 8 provide provide VB 12483 242 9 an an DT 12483 242 10 automated automate VBN 12483 242 11 verification verification NN 12483 242 12 platform platform NN 12483 242 13 for for IN 12483 242 14 all all DT 12483 242 15 news news NN 12483 242 16 content content NN 12483 242 17 existing exist VBG 12483 242 18 in in IN 12483 242 19 digital digital JJ 12483 242 20 libraries library NNS 12483 242 21 to to TO 12483 242 22 promote promote VB 12483 242 23 media medium NNS 12483 242 24 literacy literacy NN 12483 242 25 . . . 12483 243 1 ENDNOTES endnotes NN 12483 243 2 1 1 CD 12483 243 3 M. M. NNP 12483 243 4 Connor Connor NNP 12483 243 5 Sullivan Sullivan NNP 12483 243 6 , , , 12483 243 7 “ " `` 12483 243 8 Why why WRB 12483 243 9 Librarians Librarians NNPS 12483 243 10 Ca can MD 12483 243 11 n’t not RB 12483 243 12 Fight fight VB 12483 243 13 Fake Fake NNP 12483 243 14 News News NNP 12483 243 15 , , , 12483 243 16 ” " '' 12483 243 17 Journal Journal NNP 12483 243 18 of of IN 12483 243 19 Librarianship Librarianship NNP 12483 243 20 and and CC 12483 243 21 Information Information NNP 12483 243 22 Science Science NNP 12483 243 23 51 51 CD 12483 243 24 , , , 12483 243 25 no no UH 12483 243 26 . . . 12483 244 1 4 4 LS 12483 244 2 ( ( -LRB- 12483 244 3 December December NNP 12483 244 4 2019 2019 CD 12483 244 5 ) ) -RRB- 12483 244 6 : : : 12483 244 7 1146–56 1146–56 LS 12483 244 8 , , , 12483 244 9 https://doi.org/10.1177/0961000618764258 https://doi.org/10.1177/0961000618764258 UH 12483 244 10 . . . 12483 245 1 2 2 LS 12483 245 2 “ " `` 12483 245 3 Definition definition NN 12483 245 4 of of IN 12483 245 5 ' ' `` 12483 245 6 News News NNP 12483 245 7 ' ' '' 12483 245 8 , , , 12483 245 9 ” " '' 12483 245 10 available available JJ 12483 245 11 at at IN 12483 245 12 : : : 12483 245 13 https://www.collinsdictionary.com/dictionary/english/news https://www.collinsdictionary.com/dictionary/english/news NNP 12483 245 14 3 3 CD 12483 245 15 Dominic Dominic NNP 12483 245 16 DiFranzo DiFranzo NNP 12483 245 17 and and CC 12483 245 18 Kristine Kristine NNP 12483 245 19 Gloria Gloria NNP 12483 245 20 - - HYPH 12483 245 21 Garcia Garcia NNP 12483 245 22 , , , 12483 245 23 “ " `` 12483 245 24 Filter Filter NNP 12483 245 25 Bubbles Bubbles NNPS 12483 245 26 and and CC 12483 245 27 Fake Fake NNP 12483 245 28 News News NNP 12483 245 29 , , , 12483 245 30 ” " '' 12483 245 31 XRDS xrds NN 12483 245 32 : : : 12483 245 33 Crossroads Crossroads NNP 12483 245 34 , , , 12483 245 35 The the DT 12483 245 36 ACM ACM NNP 12483 245 37 Magazine Magazine NNP 12483 245 38 for for IN 12483 245 39 Students student NNS 12483 245 40 23 23 CD 12483 245 41 , , , 12483 245 42 no no UH 12483 245 43 . . . 12483 246 1 3 3 LS 12483 246 2 ( ( -LRB- 12483 246 3 April April NNP 12483 246 4 2017 2017 CD 12483 246 5 ) ) -RRB- 12483 246 6 : : : 12483 246 7 32–35 32–35 LS 12483 246 8 , , , 12483 246 9 https://doi.org/10.1145/3055153 https://doi.org/10.1145/3055153 NNP 12483 246 10 . . . 12483 247 1 https://doi.org/10.1177/0961000618764258 https://doi.org/10.1177/0961000618764258 NNP 12483 247 2 https://www.collinsdictionary.com/dictionary/english/news https://www.collinsdictionary.com/dictionary/english/news NNP 12483 247 3 https://doi.org/10.1145/3055153 https://doi.org/10.1145/3055153 NNP 12483 247 4 INFORMATION INFORMATION NNP 12483 247 5 TECHNOLOGY technology NN 12483 247 6 AND and CC 12483 247 7 LIBRARIES library NNS 12483 247 8 DECEMBER DECEMBER NNP 12483 247 9 2020 2020 CD 12483 247 10 AUTOMATED AUTOMATED NNP 12483 247 11 FAKE FAKE NNP 12483 247 12 NEWS NEWS NNP 12483 247 13 DETECTION detection VB 12483 247 14 IN in IN 12483 247 15 THE the DT 12483 247 16 AGE AGE NNP 12483 247 17 OF of IN 12483 247 18 DIGITAL DIGITAL NNP 12483 247 19 LIBRARIES library NNS 12483 247 20 | | NNP 12483 247 21 MERTOĞLU MERTOĞLU NNS 12483 247 22 AND and CC 12483 247 23 GENÇ GENÇ NNP 12483 247 24 16 16 CD 12483 247 25 4 4 CD 12483 247 26 Andrew Andrew NNP 12483 247 27 Guess Guess NNP 12483 247 28 , , , 12483 247 29 Brendan Brendan NNP 12483 247 30 Nyhan Nyhan NNP 12483 247 31 , , , 12483 247 32 and and CC 12483 247 33 Jason Jason NNP 12483 247 34 Reifler Reifler NNP 12483 247 35 , , , 12483 247 36 “ " `` 12483 247 37 Selective selective JJ 12483 247 38 Exposure exposure NN 12483 247 39 to to IN 12483 247 40 Misinformation misinformation NN 12483 247 41 : : : 12483 247 42 Evidence evidence NN 12483 247 43 from from IN 12483 247 44 the the DT 12483 247 45 Consumption Consumption NNP 12483 247 46 of of IN 12483 247 47 Fake Fake NNP 12483 247 48 News News NNP 12483 247 49 during during IN 12483 247 50 the the DT 12483 247 51 2016 2016 CD 12483 247 52 US US NNP 12483 247 53 Presidential Presidential NNP 12483 247 54 Campaign Campaign NNP 12483 247 55 , , , 12483 247 56 ” " '' 12483 247 57 European European NNP 12483 247 58 Research Research NNP 12483 247 59 Council Council NNP 12483 247 60 9 9 CD 12483 247 61 , , , 12483 247 62 no no UH 12483 247 63 . . . 12483 248 1 3 3 CD 12483 248 2 ( ( -LRB- 12483 248 3 2018 2018 CD 12483 248 4 ) ) -RRB- 12483 248 5 : : : 12483 248 6 4 4 LS 12483 248 7 ; ; : 12483 248 8 Eni Eni NNP 12483 248 9 Mustafaraj Mustafaraj NNP 12483 248 10 and and CC 12483 248 11 P. P. NNP 12483 248 12 Takis Takis NNP 12483 248 13 Metaxas Metaxas NNP 12483 248 14 , , , 12483 248 15 “ " `` 12483 248 16 The the DT 12483 248 17 Fake Fake NNP 12483 248 18 News News NNP 12483 248 19 Spreading Spreading NNP 12483 248 20 Plague Plague NNP 12483 248 21 : : : 12483 248 22 Was be VBD 12483 248 23 It -PRON- PRP 12483 248 24 Preventable preventable JJ 12483 248 25 ? ? . 12483 248 26 ” " '' 12483 248 27 Proceedings proceeding NNS 12483 248 28 of of IN 12483 248 29 the the DT 12483 248 30 2017 2017 CD 12483 248 31 ACM ACM NNP 12483 248 32 on on IN 12483 248 33 Web Web NNP 12483 248 34 Science Science NNP 12483 248 35 Conference Conference NNP 12483 248 36 , , , 12483 248 37 ( ( -LRB- 12483 248 38 June June NNP 12483 248 39 2017 2017 CD 12483 248 40 ) ) -RRB- 12483 248 41 : : : 12483 248 42 235–39 235–39 CD 12483 248 43 , , , 12483 248 44 https://doi.org/10.1145/3091478.3091523 https://doi.org/10.1145/3091478.3091523 ADD 12483 248 45 . . . 12483 249 1 5 5 CD 12483 249 2 Jana Jana NNP 12483 249 3 Laura Laura NNP 12483 249 4 Egelhofer Egelhofer NNP 12483 249 5 and and CC 12483 249 6 Sophie Sophie NNP 12483 249 7 Lecheler Lecheler NNP 12483 249 8 , , , 12483 249 9 “ " `` 12483 249 10 Fake Fake NNP 12483 249 11 News News NNP 12483 249 12 as as IN 12483 249 13 a a DT 12483 249 14 Two two CD 12483 249 15 - - HYPH 12483 249 16 Dimensional Dimensional NNP 12483 249 17 Phenomenon Phenomenon NNP 12483 249 18 : : : 12483 249 19 A a DT 12483 249 20 Framework Framework NNP 12483 249 21 and and CC 12483 249 22 Research Research NNP 12483 249 23 Agenda Agenda NNP 12483 249 24 , , , 12483 249 25 ” " '' 12483 249 26 Annals Annals NNPS 12483 249 27 of of IN 12483 249 28 the the DT 12483 249 29 International International NNP 12483 249 30 Communication Communication NNP 12483 249 31 Association Association NNP 12483 249 32 43 43 CD 12483 249 33 , , , 12483 249 34 no no UH 12483 249 35 . . . 12483 250 1 2 2 CD 12483 250 2 ( ( -LRB- 12483 250 3 2019 2019 CD 12483 250 4 ) ) -RRB- 12483 250 5 : : : 12483 250 6 97–116 97–116 CD 12483 250 7 , , , 12483 250 8 https://doi.org/10.1080/23808985.2019.1602782 https://doi.org/10.1080/23808985.2019.1602782 NNP 12483 250 9 . . . 12483 251 1 6 6 CD 12483 251 2 Hannah Hannah NNP 12483 251 3 Rashkin Rashkin NNP 12483 251 4 et et FW 12483 251 5 al al NNP 12483 251 6 . . NNP 12483 251 7 , , , 12483 251 8 “ " `` 12483 251 9 Truth Truth NNP 12483 251 10 of of IN 12483 251 11 Varying Varying NNP 12483 251 12 Shades shade NNS 12483 251 13 : : : 12483 251 14 Analyzing analyze VBG 12483 251 15 Language Language NNP 12483 251 16 in in IN 12483 251 17 Fake Fake NNP 12483 251 18 News News NNP 12483 251 19 and and CC 12483 251 20 Political Political NNP 12483 251 21 Fact Fact NNP 12483 251 22 - - HYPH 12483 251 23 Checking Checking NNP 12483 251 24 , , , 12483 251 25 ” " '' 12483 251 26 Proceedings proceeding NNS 12483 251 27 of of IN 12483 251 28 the the DT 12483 251 29 2017 2017 CD 12483 251 30 Conference Conference NNP 12483 251 31 on on IN 12483 251 32 Empirical Empirical NNP 12483 251 33 Methods Methods NNPS 12483 251 34 in in IN 12483 251 35 Natural Natural NNP 12483 251 36 Language Language NNP 12483 251 37 Processing Processing NNP 12483 251 38 , , , 12483 251 39 ( ( -LRB- 12483 251 40 2017 2017 CD 12483 251 41 ) ) -RRB- 12483 251 42 : : : 12483 251 43 2931–37 2931–37 LS 12483 251 44 . . . 12483 252 1 7 7 CD 12483 252 2 Soroush soroush JJ 12483 252 3 Vosoughi Vosoughi NNP 12483 252 4 , , , 12483 252 5 Deb Deb NNP 12483 252 6 Roy Roy NNP 12483 252 7 , , , 12483 252 8 and and CC 12483 252 9 Sinan Sinan NNP 12483 252 10 Aral Aral NNP 12483 252 11 , , , 12483 252 12 “ " `` 12483 252 13 The the DT 12483 252 14 Spread spread NN 12483 252 15 of of IN 12483 252 16 True true JJ 12483 252 17 and and CC 12483 252 18 False False NNP 12483 252 19 News News NNP 12483 252 20 Online Online NNP 12483 252 21 , , , 12483 252 22 ” " '' 12483 252 23 Science science NN 12483 252 24 359 359 CD 12483 252 25 , , , 12483 252 26 no no UH 12483 252 27 . . . 12483 253 1 6380 6380 CD 12483 253 2 ( ( -LRB- 12483 253 3 2018 2018 CD 12483 253 4 ) ) -RRB- 12483 253 5 : : : 12483 253 6 1146–51 1146–51 LS 12483 253 7 , , , 12483 253 8 https://doi.org/10.1126/science.aap9559 https://doi.org/10.1126/science.aap9559 NNP 12483 253 9 . . . 12483 254 1 8 8 CD 12483 254 2 Xinyi Xinyi NNP 12483 254 3 Zhou Zhou NNP 12483 254 4 and and CC 12483 254 5 Reza Reza NNP 12483 254 6 Zafarani Zafarani NNP 12483 254 7 , , , 12483 254 8 “ " `` 12483 254 9 A a DT 12483 254 10 Survey Survey NNP 12483 254 11 of of IN 12483 254 12 Fake Fake NNP 12483 254 13 News News NNP 12483 254 14 : : : 12483 254 15 Fundamental Fundamental NNP 12483 254 16 Theories theory NNS 12483 254 17 , , , 12483 254 18 Detection Detection NNP 12483 254 19 Methods Methods NNPS 12483 254 20 , , , 12483 254 21 and and CC 12483 254 22 Opportunities opportunity NNS 12483 254 23 , , , 12483 254 24 ” " '' 12483 254 25 ACM ACM NNP 12483 254 26 Computing Computing NNP 12483 254 27 Surveys Surveys NNP 12483 254 28 ( ( -LRB- 12483 254 29 CSUR CSUR NNP 12483 254 30 ) ) -RRB- 12483 254 31 53 53 CD 12483 254 32 , , , 12483 254 33 no no UH 12483 254 34 . . . 12483 255 1 5 5 CD 12483 255 2 ( ( -LRB- 12483 255 3 2020 2020 CD 12483 255 4 ) ) -RRB- 12483 255 5 : : : 12483 255 6 1–40 1–40 LS 12483 255 7 , , , 12483 255 8 https://doi.org/10.1145/3395046 https://doi.org/10.1145/3395046 UH 12483 255 9 . . . 12483 256 1 9 9 CD 12483 256 2 S. S. NNP 12483 256 3 F. F. NNP 12483 256 4 Kattimani Kattimani NNP 12483 256 5 , , , 12483 256 6 Praveenkumar Praveenkumar NNP 12483 256 7 Kumbargoudar Kumbargoudar NNP 12483 256 8 , , , 12483 256 9 and and CC 12483 256 10 D. D. NNP 12483 256 11 S. S. NNP 12483 256 12 Gobbur Gobbur NNP 12483 256 13 , , , 12483 256 14 “ " `` 12483 256 15 Training training NN 12483 256 16 of of IN 12483 256 17 the the DT 12483 256 18 Library Library NNP 12483 256 19 Professionals Professionals NNPS 12483 256 20 in in IN 12483 256 21 Digital Digital NNP 12483 256 22 Era Era NNP 12483 256 23 : : : 12483 256 24 Key Key NNP 12483 256 25 Issues Issues NNPS 12483 256 26 ” " '' 12483 256 27 ( ( -LRB- 12483 256 28 2006 2006 CD 12483 256 29 ) ) -RRB- 12483 256 30 , , , 12483 256 31 https://ir.inflibnet.ac.in:8443/ir/handle/1944/1234 https://ir.inflibnet.ac.in:8443/ir/handle/1944/1234 RB 12483 256 32 . . . 12483 257 1 10 10 CD 12483 257 2 Lynn Lynn NNP 12483 257 3 Silipigni Silipigni NNP 12483 257 4 Connaway Connaway NNP 12483 257 5 et et NNP 12483 257 6 al al NNP 12483 257 7 . . NNP 12483 257 8 , , , 12483 257 9 “ " `` 12483 257 10 Digital Digital NNP 12483 257 11 Literacy Literacy NNP 12483 257 12 in in IN 12483 257 13 the the DT 12483 257 14 Era Era NNP 12483 257 15 of of IN 12483 257 16 Fake Fake NNP 12483 257 17 News news NN 12483 257 18 : : : 12483 257 19 Key key JJ 12483 257 20 Roles Roles NNPS 12483 257 21 for for IN 12483 257 22 Information Information NNP 12483 257 23 Professionals Professionals NNPS 12483 257 24 , , , 12483 257 25 ” " '' 12483 257 26 Proceedings proceeding NNS 12483 257 27 of of IN 12483 257 28 the the DT 12483 257 29 Association Association NNP 12483 257 30 for for IN 12483 257 31 Information Information NNP 12483 257 32 Science Science NNP 12483 257 33 and and CC 12483 257 34 Technology Technology NNP 12483 257 35 54 54 CD 12483 257 36 , , , 12483 257 37 no no UH 12483 257 38 . . . 12483 258 1 1 1 CD 12483 258 2 ( ( -LRB- 12483 258 3 2017 2017 CD 12483 258 4 ) ) -RRB- 12483 258 5 : : : 12483 258 6 554–55 554–55 CD 12483 258 7 , , , 12483 258 8 https://doi.org/10.1002/pra2.2017.14505401070 https://doi.org/10.1002/pra2.2017.14505401070 ADD 12483 258 9 . . . 12483 259 1 11 11 CD 12483 259 2 Matthew Matthew NNP 12483 259 3 C. C. NNP 12483 259 4 Sullivan Sullivan NNP 12483 259 5 , , , 12483 259 6 “ " `` 12483 259 7 Libraries library NNS 12483 259 8 and and CC 12483 259 9 Fake Fake NNP 12483 259 10 News news NN 12483 259 11 : : : 12483 259 12 What what WP 12483 259 13 ’s ’ VBZ 12483 259 14 the the DT 12483 259 15 Problem problem NN 12483 259 16 ? ? . 12483 260 1 What what WP 12483 260 2 ’s ’ VBZ 12483 260 3 the the DT 12483 260 4 Plan Plan NNP 12483 260 5 ? ? . 12483 260 6 , , , 12483 260 7 ” " '' 12483 260 8 Communications Communications NNPS 12483 260 9 in in IN 12483 260 10 Information Information NNP 12483 260 11 Literacy Literacy NNP 12483 260 12 13 13 CD 12483 260 13 , , , 12483 260 14 no no UH 12483 260 15 . . . 12483 261 1 1 1 CD 12483 261 2 ( ( -LRB- 12483 261 3 2019 2019 CD 12483 261 4 ) ) -RRB- 12483 261 5 : : : 12483 261 6 91–113 91–113 NNP 12483 261 7 , , , 12483 261 8 https://doi.org/10.15760/comminfolit.2019.13.1.7 https://doi.org/10.15760/comminfolit.2019.13.1.7 NNP 12483 261 9 . . . 12483 262 1 12 12 CD 12483 262 2 Wayne Wayne NNP 12483 262 3 Finley Finley NNP 12483 262 4 , , , 12483 262 5 Beth Beth NNP 12483 262 6 McGowan McGowan NNP 12483 262 7 , , , 12483 262 8 and and CC 12483 262 9 Joanna Joanna NNP 12483 262 10 Kluever Kluever NNP 12483 262 11 , , , 12483 262 12 “ " `` 12483 262 13 Fake Fake NNP 12483 262 14 News News NNP 12483 262 15 : : : 12483 262 16 An an DT 12483 262 17 Opportunity opportunity NN 12483 262 18 for for IN 12483 262 19 Real real JJ 12483 262 20 Librarianship librarianship NN 12483 262 21 , , , 12483 262 22 ” " '' 12483 262 23 ILA ILA NNP 12483 262 24 reporter reporter NN 12483 262 25 35 35 CD 12483 262 26 , , , 12483 262 27 no no UH 12483 262 28 . . . 12483 263 1 3 3 CD 12483 263 2 ( ( -LRB- 12483 263 3 2017 2017 CD 12483 263 4 ) ) -RRB- 12483 263 5 : : : 12483 263 6 8–12 8–12 CD 12483 263 7 ; ; : 12483 263 8 American American NNP 12483 263 9 Library Library NNP 12483 263 10 Association Association NNP 12483 263 11 , , , 12483 263 12 “ " `` 12483 263 13 Resolution resolution NN 12483 263 14 on on IN 12483 263 15 Access Access NNP 12483 263 16 to to IN 12483 263 17 Accurate Accurate NNP 12483 263 18 Information Information NNP 12483 263 19 , , , 12483 263 20 ” " '' 12483 263 21 2018 2018 CD 12483 263 22 ; ; : 12483 263 23 Nick Nick NNP 12483 263 24 Rochlin Rochlin NNP 12483 263 25 , , , 12483 263 26 “ " `` 12483 263 27 Fake Fake NNP 12483 263 28 News News NNP 12483 263 29 : : : 12483 263 30 Belief belief NN 12483 263 31 in in IN 12483 263 32 Post Post NNP 12483 263 33 - - NNP 12483 263 34 Truth Truth NNP 12483 263 35 , , , 12483 263 36 ” " '' 12483 263 37 Library library JJ 12483 263 38 Hi Hi NNP 12483 263 39 Tech Tech NNP 12483 263 40 35 35 CD 12483 263 41 , , , 12483 263 42 no no UH 12483 263 43 . . . 12483 264 1 3 3 CD 12483 264 2 ( ( -LRB- 12483 264 3 2017 2017 CD 12483 264 4 ) ) -RRB- 12483 264 5 : : : 12483 264 6 386–92 386–92 CD 12483 264 7 , , , 12483 264 8 https://doi.org/10.1108/LHT-03-2017-0062 https://doi.org/10.1108/LHT-03-2017-0062 NNS 12483 264 9 ; ; : 12483 264 10 Linda Linda NNP 12483 264 11 Jacobson Jacobson NNP 12483 264 12 , , , 12483 264 13 “ " `` 12483 264 14 The the DT 12483 264 15 Smell Smell NNP 12483 264 16 Test test NN 12483 264 17 : : : 12483 264 18 In in IN 12483 264 19 the the DT 12483 264 20 Era Era NNP 12483 264 21 of of IN 12483 264 22 Fake Fake NNP 12483 264 23 News News NNP 12483 264 24 , , , 12483 264 25 Librarians Librarians NNPS 12483 264 26 Are be VBP 12483 264 27 Our -PRON- PRP$ 12483 264 28 Best good JJS 12483 264 29 Hope hope NN 12483 264 30 , , , 12483 264 31 ” " '' 12483 264 32 School School NNP 12483 264 33 Library Library NNP 12483 264 34 Journal Journal NNP 12483 264 35 63 63 CD 12483 264 36 , , , 12483 264 37 no no UH 12483 264 38 . . . 12483 265 1 1 1 CD 12483 265 2 ( ( -LRB- 12483 265 3 2017 2017 CD 12483 265 4 ) ) -RRB- 12483 265 5 : : : 12483 265 6 24–29 24–29 LS 12483 265 7 ; ; : 12483 265 8 Angeleen Angeleen NNP 12483 265 9 Neely neely RB 12483 265 10 – – : 12483 265 11 Sardon Sardon NNP 12483 265 12 , , , 12483 265 13 and and CC 12483 265 14 Mia Mia NNP 12483 265 15 Tignor Tignor NNP 12483 265 16 , , , 12483 265 17 “ " `` 12483 265 18 Focus focus VB 12483 265 19 on on IN 12483 265 20 the the DT 12483 265 21 Facts fact NNS 12483 265 22 : : : 12483 265 23 A A NNP 12483 265 24 News News NNP 12483 265 25 and and CC 12483 265 26 Information Information NNP 12483 265 27 Literacy Literacy NNP 12483 265 28 Instructional Instructional NNP 12483 265 29 Program Program NNP 12483 265 30 , , , 12483 265 31 ” " '' 12483 265 32 The the DT 12483 265 33 Reference Reference NNP 12483 265 34 Librarian Librarian NNP 12483 265 35 59 59 CD 12483 265 36 , , , 12483 265 37 no no UH 12483 265 38 . . . 12483 266 1 3 3 CD 12483 266 2 ( ( -LRB- 12483 266 3 2018 2018 CD 12483 266 4 ) ) -RRB- 12483 266 5 : : : 12483 266 6 108–21 108–21 CD 12483 266 7 , , , 12483 266 8 https://doi.org https://doi.org CD 12483 266 9 /10.1080/02763877.2018.1468849 /10.1080/02763877.2018.1468849 . 12483 266 10 ; ; : 12483 266 11 Claire Claire NNP 12483 266 12 Wardle Wardle NNP 12483 266 13 and and CC 12483 266 14 Hossein Hossein NNP 12483 266 15 Derakhshan Derakhshan NNP 12483 266 16 , , , 12483 266 17 “ " `` 12483 266 18 Information Information NNP 12483 266 19 Disorder Disorder NNP 12483 266 20 : : : 12483 266 21 Toward toward IN 12483 266 22 an an DT 12483 266 23 Interdisciplinary Interdisciplinary NNP 12483 266 24 Framework Framework NNP 12483 266 25 for for IN 12483 266 26 Research Research NNP 12483 266 27 and and CC 12483 266 28 Policy Policy NNP 12483 266 29 Making Making NNP 12483 266 30 , , , 12483 266 31 ” " '' 12483 266 32 Council Council NNP 12483 266 33 of of IN 12483 266 34 Europe Europe NNP 12483 266 35 report report VBP 12483 266 36 27 27 CD 12483 266 37 ( ( -LRB- 12483 266 38 2017 2017 CD 12483 266 39 ) ) -RRB- 12483 266 40 . . . 12483 267 1 13 13 CD 12483 267 2 IFLA ifla NN 12483 267 3 , , , 12483 267 4 “ " `` 12483 267 5 How how WRB 12483 267 6 to to TO 12483 267 7 Spot Spot NNP 12483 267 8 Fake Fake NNP 12483 267 9 News News NNP 12483 267 10 , , , 12483 267 11 ” " '' 12483 267 12 2017 2017 CD 12483 267 13 . . . 12483 267 14 https://doi.org/10.1145/3091478.3091523 https://doi.org/10.1145/3091478.3091523 ADD 12483 267 15 https://doi.org/10.1080/23808985.2019.1602782 https://doi.org/10.1080/23808985.2019.1602782 NN 12483 267 16 https://doi.org/10.1145/3395046 https://doi.org/10.1145/3395046 UH 12483 267 17 https://doi.org/10.1002/pra2.2017.14505401070 https://doi.org/10.1002/pra2.2017.14505401070 VBZ 12483 267 18 https://doi.org/10.15760/comminfolit.2019.13.1.7 https://doi.org/10.15760/comminfolit.2019.13.1.7 JJ 12483 267 19 https://www.emerald.com/insight/publication/issn/0737-8831 https://www.emerald.com/insight/publication/issn/0737-8831 JJ 12483 267 20 https://doi.org/10.1108/LHT-03-2017-0062 https://doi.org/10.1108/LHT-03-2017-0062 NNS 12483 267 21 INFORMATION INFORMATION VBD 12483 267 22 TECHNOLOGY technology NN 12483 267 23 AND and CC 12483 267 24 LIBRARIES library NNS 12483 267 25 DECEMBER DECEMBER NNP 12483 267 26 2020 2020 CD 12483 267 27 AUTOMATED AUTOMATED NNP 12483 267 28 FAKE FAKE NNP 12483 267 29 NEWS NEWS NNP 12483 267 30 DETECTION detection VB 12483 267 31 IN in IN 12483 267 32 THE the DT 12483 267 33 AGE AGE NNP 12483 267 34 OF of IN 12483 267 35 DIGITAL DIGITAL NNP 12483 267 36 LIBRARIES library NNS 12483 267 37 | | NNP 12483 267 38 MERTOĞLU MERTOĞLU NNS 12483 267 39 AND and CC 12483 267 40 GENÇ genç VBP 12483 267 41 17 17 CD 12483 267 42 14 14 CD 12483 267 43 Jane Jane NNP 12483 267 44 Mandalios Mandalios NNPS 12483 267 45 , , , 12483 267 46 “ " `` 12483 267 47 Radar radar NN 12483 267 48 : : : 12483 267 49 An an DT 12483 267 50 Approach Approach NNP 12483 267 51 for for IN 12483 267 52 Helping help VBG 12483 267 53 Students Students NNP 12483 267 54 Evaluate Evaluate NNP 12483 267 55 Internet internet NN 12483 267 56 Sources source NNS 12483 267 57 , , , 12483 267 58 ” " '' 12483 267 59 Journal Journal NNP 12483 267 60 of of IN 12483 267 61 Information Information NNP 12483 267 62 Science Science NNP 12483 267 63 39 39 CD 12483 267 64 , , , 12483 267 65 no no UH 12483 267 66 . . . 12483 268 1 4 4 CD 12483 268 2 ( ( -LRB- 12483 268 3 2013 2013 CD 12483 268 4 ) ) -RRB- 12483 268 5 : : : 12483 268 6 470–78 470–78 CD 12483 268 7 , , , 12483 268 8 https://doi.org/10.1177/0165551513478889 https://doi.org/10.1177/0165551513478889 PRP 12483 268 9 ; ; : 12483 268 10 Sarah Sarah NNP 12483 268 11 Blakeslee Blakeslee NNP 12483 268 12 , , , 12483 268 13 “ " `` 12483 268 14 The the DT 12483 268 15 CRAAP CRAAP NNP 12483 268 16 test test NN 12483 268 17 , , , 12483 268 18 ” " '' 12483 268 19 LOEX LOEX NNP 12483 268 20 Quarterly Quarterly NNP 12483 268 21 3 3 CD 12483 268 22 , , , 12483 268 23 no no UH 12483 268 24 . . . 12483 269 1 3 3 CD 12483 269 2 ( ( -LRB- 12483 269 3 2004):4 2004):4 CD 12483 269 4 . . . 12483 270 1 15 15 CD 12483 270 2 Victoria Victoria NNP 12483 270 3 L. L. NNP 12483 270 4 Rubin Rubin NNP 12483 270 5 and and CC 12483 270 6 Niall Niall NNP 12483 270 7 Conroy Conroy NNP 12483 270 8 , , , 12483 270 9 “ " `` 12483 270 10 Discerning Discerning NNP 12483 270 11 Truth Truth NNP 12483 270 12 from from IN 12483 270 13 Deception deception NN 12483 270 14 : : : 12483 270 15 Human Human NNP 12483 270 16 Judgments Judgments NNPS 12483 270 17 and and CC 12483 270 18 Automation Automation NNP 12483 270 19 Efforts Efforts NNPS 12483 270 20 , , , 12483 270 21 ” " '' 12483 270 22 First First NNP 12483 270 23 Monday Monday NNP 12483 270 24 17 17 CD 12483 270 25 , , , 12483 270 26 no no UH 12483 270 27 . . . 12483 271 1 5 5 CD 12483 271 2 ( ( -LRB- 12483 271 3 2012 2012 CD 12483 271 4 ) ) -RRB- 12483 271 5 , , , 12483 271 6 https://doi.org/10.5210/fm.v17i3.3933 https://doi.org/10.5210/fm.v17i3.3933 LS 12483 271 7 ; ; : 12483 271 8 Verónica Verónica NNP 12483 271 9 Pérez Pérez NNP 12483 271 10 - - HYPH 12483 271 11 Rosas Rosas NNP 12483 271 12 et et NNP 12483 271 13 al al NNP 12483 271 14 . . NNP 12483 271 15 , , , 12483 271 16 “ " `` 12483 271 17 Automatic automatic JJ 12483 271 18 Detection detection NN 12483 271 19 of of IN 12483 271 20 Fake Fake NNP 12483 271 21 News News NNP 12483 271 22 , , , 12483 271 23 ” " '' 12483 271 24 arXiv arXiv NNP 12483 271 25 preprint preprint NN 12483 271 26 arXiv:1708.07104 arXiv:1708.07104 NNP 12483 271 27 ( ( -LRB- 12483 271 28 2017 2017 CD 12483 271 29 ) ) -RRB- 12483 271 30 . . . 12483 272 1 16 16 CD 12483 272 2 Justin Justin NNP 12483 272 3 P. P. NNP 12483 272 4 Friesen Friesen NNP 12483 272 5 , , , 12483 272 6 Troy Troy NNP 12483 272 7 H. H. NNP 12483 272 8 Campbell Campbell NNP 12483 272 9 , , , 12483 272 10 and and CC 12483 272 11 Aaron Aaron NNP 12483 272 12 C. C. NNP 12483 272 13 Kay Kay NNP 12483 272 14 , , , 12483 272 15 “ " `` 12483 272 16 The the DT 12483 272 17 Psychological Psychological NNP 12483 272 18 Advantage Advantage NNP 12483 272 19 of of IN 12483 272 20 Unfalsifiability Unfalsifiability NNP 12483 272 21 : : : 12483 272 22 The the DT 12483 272 23 Appeal appeal NN 12483 272 24 of of IN 12483 272 25 Untestable untestable JJ 12483 272 26 Religious Religious NNP 12483 272 27 and and CC 12483 272 28 Political Political NNP 12483 272 29 Ideologies Ideologies NNPS 12483 272 30 , , , 12483 272 31 ” " '' 12483 272 32 Journal Journal NNP 12483 272 33 of of IN 12483 272 34 Personality Personality NNP 12483 272 35 and and CC 12483 272 36 Social Social NNP 12483 272 37 Psychology Psychology NNP 12483 272 38 108 108 CD 12483 272 39 , , , 12483 272 40 no no UH 12483 272 41 . . . 12483 273 1 3 3 CD 12483 273 2 ( ( -LRB- 12483 273 3 2015 2015 CD 12483 273 4 ) ) -RRB- 12483 273 5 : : : 12483 273 6 515–29 515–29 CD 12483 273 7 , , , 12483 273 8 https://doi.org/10.1037/pspp0000018 https://doi.org/10.1037/pspp0000018 ADD 12483 273 9 . . . 12483 274 1 17 17 CD 12483 274 2 Tanja Tanja NNP 12483 274 3 Pavleska Pavleska NNP 12483 274 4 et et NNP 12483 274 5 al al NNP 12483 274 6 . . NNP 12483 274 7 , , , 12483 274 8 “ " `` 12483 274 9 Performance Performance NNP 12483 274 10 Analysis Analysis NNP 12483 274 11 of of IN 12483 274 12 Fact Fact NNP 12483 274 13 - - HYPH 12483 274 14 Checking Checking NNP 12483 274 15 Organizations Organizations NNPS 12483 274 16 and and CC 12483 274 17 Initiatives Initiatives NNPS 12483 274 18 in in IN 12483 274 19 Europe Europe NNP 12483 274 20 : : : 12483 274 21 A a DT 12483 274 22 Critical Critical NNP 12483 274 23 Overview Overview NNP 12483 274 24 of of IN 12483 274 25 Online Online NNP 12483 274 26 Platforms Platforms NNP 12483 274 27 Fighting Fighting NNP 12483 274 28 Fake Fake NNP 12483 274 29 News news NN 12483 274 30 , , , 12483 274 31 ” " '' 12483 274 32 Social Social NNP 12483 274 33 Media Media NNP 12483 274 34 and and CC 12483 274 35 Convergence Convergence NNP 12483 274 36 29 29 CD 12483 274 37 ( ( -LRB- 12483 274 38 2018 2018 CD 12483 274 39 ) ) -RRB- 12483 274 40 . . . 12483 275 1 18 18 CD 12483 275 2 Yasmine Yasmine NNP 12483 275 3 Lahlou Lahlou NNP 12483 275 4 , , , 12483 275 5 Sanaa Sanaa NNP 12483 275 6 El El NNP 12483 275 7 Fkihi Fkihi NNP 12483 275 8 , , , 12483 275 9 and and CC 12483 275 10 Rdouan Rdouan NNP 12483 275 11 Faizi Faizi NNP 12483 275 12 , , , 12483 275 13 “ " `` 12483 275 14 Automatic automatic JJ 12483 275 15 Detection detection NN 12483 275 16 of of IN 12483 275 17 Fake Fake NNP 12483 275 18 News News NNP 12483 275 19 on on IN 12483 275 20 Online Online NNP 12483 275 21 Platforms Platforms NNP 12483 275 22 : : : 12483 275 23 A A NNP 12483 275 24 Survey Survey NNP 12483 275 25 , , , 12483 275 26 ” " '' 12483 275 27 ( ( -LRB- 12483 275 28 paper paper NN 12483 275 29 , , , 12483 275 30 2019 2019 CD 12483 275 31 1st 1st NN 12483 275 32 International International NNP 12483 275 33 Conference Conference NNP 12483 275 34 on on IN 12483 275 35 Smart Smart NNP 12483 275 36 Systems Systems NNPS 12483 275 37 and and CC 12483 275 38 Data Data NNPS 12483 275 39 Science Science NNP 12483 275 40 ( ( -LRB- 12483 275 41 ICSSD ICSSD NNP 12483 275 42 ) ) -RRB- 12483 275 43 , , , 12483 275 44 Rabat Rabat NNP 12483 275 45 , , , 12483 275 46 Morocco Morocco NNP 12483 275 47 , , , 12483 275 48 2019 2019 CD 12483 275 49 ) ) -RRB- 12483 275 50 , , , 12483 275 51 https://doi.org/10.1109/ICSSD47982.2019.9002823 https://doi.org/10.1109/ICSSD47982.2019.9002823 NNP 12483 275 52 ; ; : 12483 275 53 Christian Christian NNP 12483 275 54 Janze Janze NNP 12483 275 55 , , , 12483 275 56 and and CC 12483 275 57 Marten Marten NNP 12483 275 58 Risius Risius NNP 12483 275 59 , , , 12483 275 60 “ " `` 12483 275 61 Automatic automatic JJ 12483 275 62 Detection detection NN 12483 275 63 of of IN 12483 275 64 Fake Fake NNP 12483 275 65 News News NNP 12483 275 66 on on IN 12483 275 67 Social Social NNP 12483 275 68 Media Media NNPS 12483 275 69 Platforms Platforms NNP 12483 275 70 , , , 12483 275 71 ” " '' 12483 275 72 ( ( -LRB- 12483 275 73 paper paper NN 12483 275 74 , , , 12483 275 75 Pasific Pasific NNP 12483 275 76 Asia Asia NNP 12483 275 77 Conference Conference NNP 12483 275 78 on on IN 12483 275 79 Information Information NNP 12483 275 80 Systems Systems NNP 12483 275 81 ( ( -LRB- 12483 275 82 PACIS PACIS NNP 12483 275 83 ) ) -RRB- 12483 275 84 , , , 12483 275 85 2017 2017 CD 12483 275 86 ) ) -RRB- 12483 275 87 ; ; : 12483 275 88 Torstein Torstein NNP 12483 275 89 Granskogen Granskogen NNP 12483 275 90 , , , 12483 275 91 “ " `` 12483 275 92 Automatic automatic JJ 12483 275 93 Detection detection NN 12483 275 94 of of IN 12483 275 95 Fake Fake NNP 12483 275 96 News News NNP 12483 275 97 in in IN 12483 275 98 Social Social NNP 12483 275 99 Media medium NNS 12483 275 100 Using use VBG 12483 275 101 Contextual Contextual NNP 12483 275 102 Information Information NNP 12483 275 103 ” " '' 12483 275 104 ( ( -LRB- 12483 275 105 master master NN 12483 275 106 ’s ’s POS 12483 275 107 thesis thesis NN 12483 275 108 , , , 12483 275 109 Norwegian Norwegian NNP 12483 275 110 University University NNP 12483 275 111 of of IN 12483 275 112 Science Science NNP 12483 275 113 and and CC 12483 275 114 Technology Technology NNP 12483 275 115 ( ( -LRB- 12483 275 116 NTNU NTNU NNP 12483 275 117 ) ) -RRB- 12483 275 118 , , , 12483 275 119 2018 2018 CD 12483 275 120 ) ) -RRB- 12483 275 121 . . . 12483 276 1 19 19 CD 12483 276 2 Jacob Jacob NNP 12483 276 3 L. L. NNP 12483 276 4 Nelson Nelson NNP 12483 276 5 and and CC 12483 276 6 Harsh Harsh NNP 12483 276 7 Taneja Taneja NNPS 12483 276 8 , , , 12483 276 9 “ " `` 12483 276 10 The the DT 12483 276 11 Small small JJ 12483 276 12 , , , 12483 276 13 Disloyal Disloyal NNP 12483 276 14 Fake Fake NNP 12483 276 15 News News NNP 12483 276 16 Audience audience NN 12483 276 17 : : : 12483 276 18 The the DT 12483 276 19 Role role NN 12483 276 20 of of IN 12483 276 21 Audience Audience NNP 12483 276 22 Availability Availability NNP 12483 276 23 in in IN 12483 276 24 Fake Fake NNP 12483 276 25 News News NNP 12483 276 26 Consumption Consumption NNP 12483 276 27 , , , 12483 276 28 ” " '' 12483 276 29 New New NNP 12483 276 30 Media Media NNP 12483 276 31 & & CC 12483 276 32 Society Society NNP 12483 276 33 20 20 CD 12483 276 34 , , , 12483 276 35 no no UH 12483 276 36 . . . 12483 277 1 10 10 CD 12483 277 2 ( ( -LRB- 12483 277 3 2018 2018 CD 12483 277 4 ) ) -RRB- 12483 277 5 : : : 12483 277 6 3720–37 3720–37 LS 12483 277 7 , , , 12483 277 8 https://doi.org/10.1177/1461444818758715 https://doi.org/10.1177/1461444818758715 JJ 12483 277 9 ; ; : 12483 277 10 Philip Philip NNP 12483 277 11 N. N. NNP 12483 277 12 Howard Howard NNP 12483 277 13 et et NNP 12483 277 14 al al NNP 12483 277 15 . . NNP 12483 277 16 , , , 12483 277 17 “ " `` 12483 277 18 Social Social NNP 12483 277 19 Media Media NNP 12483 277 20 , , , 12483 277 21 News News NNP 12483 277 22 and and CC 12483 277 23 Political Political NNP 12483 277 24 Information Information NNP 12483 277 25 During during IN 12483 277 26 the the DT 12483 277 27 US US NNP 12483 277 28 Election Election NNP 12483 277 29 : : : 12483 277 30 Was be VBD 12483 277 31 Polarizing polarize VBG 12483 277 32 Content content NN 12483 277 33 Concentrated Concentrated NNP 12483 277 34 in in IN 12483 277 35 Swing Swing NNP 12483 277 36 States States NNP 12483 277 37 ? ? . 12483 277 38 , , , 12483 277 39 ” " '' 12483 277 40 arXiv arXiv NNP 12483 277 41 preprint preprint NN 12483 277 42 arXiv:1802.03573 arXiv:1802.03573 NNP 12483 277 43 ( ( -LRB- 12483 277 44 2018 2018 CD 12483 277 45 ) ) -RRB- 12483 277 46 ; ; : 12483 277 47 Alexandre Alexandre NNP 12483 277 48 Bovet Bovet NNPS 12483 277 49 and and CC 12483 277 50 Hernán Hernán NNP 12483 277 51 A. A. NNP 12483 277 52 Makse Makse NNP 12483 277 53 , , , 12483 277 54 “ " `` 12483 277 55 Influence Influence NNP 12483 277 56 of of IN 12483 277 57 Fake Fake NNP 12483 277 58 News News NNP 12483 277 59 in in IN 12483 277 60 Twitter Twitter NNP 12483 277 61 During during IN 12483 277 62 the the DT 12483 277 63 2016 2016 CD 12483 277 64 US US NNP 12483 277 65 Presidential Presidential NNP 12483 277 66 Election Election NNP 12483 277 67 , , , 12483 277 68 ” " '' 12483 277 69 Nature Nature NNP 12483 277 70 Communications Communications NNPS 12483 277 71 10 10 CD 12483 277 72 , , , 12483 277 73 no no UH 12483 277 74 . . . 12483 278 1 7 7 CD 12483 278 2 ( ( -LRB- 12483 278 3 2019 2019 CD 12483 278 4 ) ) -RRB- 12483 278 5 : : : 12483 278 6 1–14 1–14 CD 12483 278 7 , , , 12483 278 8 https://doi.org/10.1038/s41467-018-07761-2 https://doi.org/10.1038/s41467-018-07761-2 CD 12483 278 9 . . . 12483 279 1 20 20 CD 12483 279 2 Lina Lina NNP 12483 279 3 Zhou Zhou NNP 12483 279 4 et et NNP 12483 279 5 al al NNP 12483 279 6 . . NNP 12483 279 7 , , , 12483 279 8 “ " `` 12483 279 9 Automating Automating NNP 12483 279 10 Linguistics Linguistics NNP 12483 279 11 - - HYPH 12483 279 12 Based base VBN 12483 279 13 Cues Cues NNP 12483 279 14 for for IN 12483 279 15 Detecting detect VBG 12483 279 16 Deception Deception NNP 12483 279 17 in in IN 12483 279 18 Text Text NNP 12483 279 19 - - HYPH 12483 279 20 Based base VBN 12483 279 21 Asynchronous Asynchronous NNP 12483 279 22 Computer Computer NNP 12483 279 23 - - HYPH 12483 279 24 Mediated Mediated NNP 12483 279 25 Communications Communications NNP 12483 279 26 , , , 12483 279 27 ” " '' 12483 279 28 Group Group NNP 12483 279 29 Decision Decision NNP 12483 279 30 and and CC 12483 279 31 Negotiation Negotiation NNP 12483 279 32 13 13 CD 12483 279 33 , , , 12483 279 34 no no UH 12483 279 35 . . . 12483 280 1 1 1 CD 12483 280 2 ( ( -LRB- 12483 280 3 2004 2004 CD 12483 280 4 ) ) -RRB- 12483 280 5 : : : 12483 280 6 81–106 81–106 CD 12483 280 7 , , , 12483 280 8 https://doi.org/10.1023/B:GRUP.0000011944.62889.6f https://doi.org/10.1023/b:grup.0000011944.62889.6f CD 12483 280 9 ; ; : 12483 280 10 Myle Myle NNP 12483 280 11 Ott Ott NNP 12483 280 12 et et FW 12483 280 13 al al NNP 12483 280 14 . . NNP 12483 280 15 , , , 12483 280 16 “ " `` 12483 280 17 Finding Finding NNP 12483 280 18 Deceptive deceptive JJ 12483 280 19 Opinion opinion NN 12483 280 20 Spam Spam NNP 12483 280 21 by by IN 12483 280 22 Any any DT 12483 280 23 Stretch Stretch NNP 12483 280 24 of of IN 12483 280 25 the the DT 12483 280 26 Imagination Imagination NNP 12483 280 27 , , , 12483 280 28 ” " '' 12483 280 29 arXiv arXiv NNP 12483 280 30 preprint preprint NN 12483 280 31 arXiv:1107.4557 arxiv:1107.4557 NN 12483 280 32 ( ( -LRB- 12483 280 33 2011 2011 CD 12483 280 34 ) ) -RRB- 12483 280 35 ; ; : 12483 280 36 Rada Rada NNP 12483 280 37 Mihalcea Mihalcea NNP 12483 280 38 and and CC 12483 280 39 Carlo Carlo NNP 12483 280 40 Strapparava Strapparava NNP 12483 280 41 , , , 12483 280 42 “ " `` 12483 280 43 The the DT 12483 280 44 Lie Lie NNP 12483 280 45 Detector detector NN 12483 280 46 : : : 12483 280 47 Explorations exploration NNS 12483 280 48 in in IN 12483 280 49 the the DT 12483 280 50 Automatic Automatic NNP 12483 280 51 Recognition Recognition NNP 12483 280 52 of of IN 12483 280 53 Deceptive Deceptive NNP 12483 280 54 Language Language NNP 12483 280 55 , , , 12483 280 56 ” " '' 12483 280 57 ( ( -LRB- 12483 280 58 paper paper NN 12483 280 59 , , , 12483 280 60 Proceedings Proceedings NNPS 12483 280 61 of of IN 12483 280 62 the the DT 12483 280 63 ACL ACL NNP 12483 280 64 - - HYPH 12483 280 65 IJCNLP IJCNLP NNP 12483 280 66 2009 2009 CD 12483 280 67 Conference Conference NNP 12483 280 68 Short Short NNP 12483 280 69 Papers Papers NNPS 12483 280 70 , , , 12483 280 71 ( ( -LRB- 12483 280 72 2009 2009 CD 12483 280 73 ) ) -RRB- 12483 280 74 : : : 12483 280 75 Association Association NNP 12483 280 76 for for IN 12483 280 77 Computational Computational NNP 12483 280 78 Linguistics Linguistics NNPS 12483 280 79 , , , 12483 280 80 309–12 309–12 CD 12483 280 81 ) ) -RRB- 12483 280 82 ; ; : 12483 280 83 Julia Julia NNP 12483 280 84 B. B. NNP 12483 280 85 Hirschberg Hirschberg NNP 12483 280 86 et et NNP 12483 280 87 al al NNP 12483 280 88 . . NNP 12483 280 89 , , , 12483 280 90 “ " `` 12483 280 91 Distinguishing distinguish VBG 12483 280 92 Deceptive Deceptive NNP 12483 280 93 from from IN 12483 280 94 Non non JJ 12483 280 95 - - JJ 12483 280 96 Deceptive deceptive JJ 12483 280 97 Speech speech NN 12483 280 98 , , , 12483 280 99 ” " '' 12483 280 100 ( ( -LRB- 12483 280 101 2005 2005 CD 12483 280 102 ) ) -RRB- 12483 280 103 , , , 12483 280 104 https://doi.org/10.7916/D8697C06 https://doi.org/10.7916/D8697C06 NNP 12483 280 105 . . . 12483 281 1 21 21 CD 12483 281 2 Victoria Victoria NNP 12483 281 3 L. L. NNP 12483 281 4 Rubin Rubin NNP 12483 281 5 , , , 12483 281 6 Yimin Yimin NNP 12483 281 7 Chen Chen NNP 12483 281 8 , , , 12483 281 9 and and CC 12483 281 10 Nadia Nadia NNP 12483 281 11 K. K. NNP 12483 281 12 Conroy Conroy NNP 12483 281 13 , , , 12483 281 14 “ " `` 12483 281 15 Deception deception NN 12483 281 16 Detection detection NN 12483 281 17 for for IN 12483 281 18 News news NN 12483 281 19 : : : 12483 281 20 Three three CD 12483 281 21 Types Types NNPS 12483 281 22 of of IN 12483 281 23 Fakes fake NNS 12483 281 24 , , , 12483 281 25 ” " '' 12483 281 26 Proceedings proceeding NNS 12483 281 27 of of IN 12483 281 28 the the DT 12483 281 29 Association Association NNP 12483 281 30 for for IN 12483 281 31 Information Information NNP 12483 281 32 Science Science NNP 12483 281 33 and and CC 12483 281 34 Technology Technology NNP 12483 281 35 52 52 CD 12483 281 36 , , , 12483 281 37 no no UH 12483 281 38 . . . 12483 282 1 1 1 CD 12483 282 2 ( ( -LRB- 12483 282 3 2015 2015 CD 12483 282 4 ) ) -RRB- 12483 282 5 : : : 12483 282 6 1–4 1–4 CD 12483 282 7 , , , 12483 282 8 https://doi.org/10.1002/pra2.2015.145052010083 https://doi.org/10.1002/pra2.2015.145052010083 ADD 12483 282 9 ; ; : 12483 282 10 David David NNP 12483 282 11 M. M. NNP 12483 282 12 Markowitz Markowitz NNP 12483 282 13 , , , 12483 282 14 and and CC 12483 282 15 Jeffrey Jeffrey NNP 12483 282 16 T. T. NNP 12483 282 17 Hancock Hancock NNP 12483 282 18 , , , 12483 282 19 “ " `` 12483 282 20 Linguistic linguistic JJ 12483 282 21 Traces trace NNS 12483 282 22 of of IN 12483 282 23 a a DT 12483 282 24 Scientific Scientific NNP 12483 282 25 Fraud Fraud NNP 12483 282 26 : : : 12483 282 27 The the DT 12483 282 28 Case Case NNP 12483 282 29 of of IN 12483 282 30 Diederik Diederik NNP 12483 282 31 Stapel Stapel NNP 12483 282 32 , , , 12483 282 33 ” " '' 12483 282 34 PloS plos NN 12483 282 35 https://doi.org/10.1177/0165551513478889 https://doi.org/10.1177/0165551513478889 PRP$ 12483 282 36 https://doi.org/10.5210/fm.v17i3.3933 https://doi.org/10.5210/fm.v17i3.3933 ADD 12483 282 37 https://psycnet.apa.org/doi/10.1037/pspp0000018 https://psycnet.apa.org/doi/10.1037/pspp0000018 . 12483 282 38 https://doi.org/10.1109/ICSSD47982.2019.9002823 https://doi.org/10.1109/ICSSD47982.2019.9002823 NNP 12483 282 39 https://doi.org/10.1177%2F1461444818758715 https://doi.org/10.1177%2F1461444818758715 NNP 12483 282 40 https://doi.org/10.1038/s41467-018-07761-2 https://doi.org/10.1038/s41467-018-07761-2 CD 12483 282 41 https://doi.org/10.1023/B:GRUP.0000011944.62889.6f https://doi.org/10.1023/b:grup.0000011944.62889.6f CD 12483 282 42 https://doi.org/10.7916/D8697C06 https://doi.org/10.7916/D8697C06 NNP 12483 282 43 https://doi.org/10.1002/pra2.2015.145052010083 https://doi.org/10.1002/pra2.2015.145052010083 '' 12483 282 44 INFORMATION information NN 12483 282 45 TECHNOLOGY technology NN 12483 282 46 AND and CC 12483 282 47 LIBRARIES library NNS 12483 282 48 DECEMBER DECEMBER NNP 12483 282 49 2020 2020 CD 12483 282 50 AUTOMATED AUTOMATED NNP 12483 282 51 FAKE FAKE NNP 12483 282 52 NEWS NEWS NNP 12483 282 53 DETECTION detection VB 12483 282 54 IN in IN 12483 282 55 THE the DT 12483 282 56 AGE AGE NNP 12483 282 57 OF of IN 12483 282 58 DIGITAL DIGITAL NNP 12483 282 59 LIBRARIES library NNS 12483 282 60 | | NNP 12483 282 61 MERTOĞLU MERTOĞLU NNS 12483 282 62 AND and CC 12483 282 63 GENÇ genç VBP 12483 282 64 18 18 CD 12483 282 65 one one CD 12483 282 66 9 9 CD 12483 282 67 , , , 12483 282 68 no no UH 12483 282 69 . . . 12483 283 1 8 8 CD 12483 283 2 ( ( -LRB- 12483 283 3 2014 2014 CD 12483 283 4 ) ) -RRB- 12483 283 5 : : : 12483 283 6 e105937 e105937 NN 12483 283 7 , , , 12483 283 8 https://doi.org/10.1371/journal.pone.0105937 https://doi.org/10.1371/journal.pone.0105937 NN 12483 283 9 ; ; : 12483 283 10 Jing Jing NNP 12483 283 11 Ma Ma NNP 12483 283 12 et et FW 12483 283 13 al al NNP 12483 283 14 . . NNP 12483 283 15 , , , 12483 283 16 “ " `` 12483 283 17 Detecting detect VBG 12483 283 18 Rumors Rumors NNPS 12483 283 19 from from IN 12483 283 20 Microblogs Microblogs NNP 12483 283 21 with with IN 12483 283 22 Recurrent Recurrent NNP 12483 283 23 Neural Neural NNP 12483 283 24 Networks Networks NNPS 12483 283 25 , , , 12483 283 26 ” " '' 12483 283 27 ( ( -LRB- 12483 283 28 paper paper NN 12483 283 29 , , , 12483 283 30 Proceedings Proceedings NNPS 12483 283 31 of of IN 12483 283 32 the the DT 12483 283 33 25th 25th JJ 12483 283 34 International International NNP 12483 283 35 Joint Joint NNP 12483 283 36 Conference Conference NNP 12483 283 37 on on IN 12483 283 38 Artificial Artificial NNP 12483 283 39 Intelligence Intelligence NNP 12483 283 40 ( ( -LRB- 12483 283 41 IJCAI IJCAI NNP 12483 283 42 2016 2016 CD 12483 283 43 ) ) -RRB- 12483 283 44 , , , 12483 283 45 ( ( -LRB- 12483 283 46 2016 2016 CD 12483 283 47 ) ) -RRB- 12483 283 48 : : : 12483 283 49 3818–24 3818–24 LS 12483 283 50 ) ) -RRB- 12483 283 51 , , , 12483 283 52 https://ink.library.smu.edu.sg/sis_research/4630 https://ink.library.smu.edu.sg/sis_research/4630 ADD 12483 283 53 . . . 12483 284 1 22 22 CD 12483 284 2 Kai Kai NNP 12483 284 3 Shu Shu NNP 12483 284 4 et et NNP 12483 284 5 al al NNP 12483 284 6 . . NNP 12483 284 7 , , , 12483 284 8 “ " `` 12483 284 9 Fake Fake NNP 12483 284 10 News News NNP 12483 284 11 Detection detection NN 12483 284 12 on on IN 12483 284 13 Social Social NNP 12483 284 14 Media medium NNS 12483 284 15 : : : 12483 284 16 A A NNP 12483 284 17 Data Data NNP 12483 284 18 Mining Mining NNP 12483 284 19 Perspective Perspective NNP 12483 284 20 , , , 12483 284 21 ” " '' 12483 284 22 ACM ACM NNP 12483 284 23 SIGKDD SIGKDD NNP 12483 284 24 Explorations Explorations NNPS 12483 284 25 Newsletter Newsletter NNP 12483 284 26 19 19 CD 12483 284 27 , , , 12483 284 28 no no UH 12483 284 29 . . . 12483 285 1 1 1 CD 12483 285 2 ( ( -LRB- 12483 285 3 2017 2017 CD 12483 285 4 ) ) -RRB- 12483 285 5 : : : 12483 285 6 22–36 22–36 CD 12483 285 7 , , , 12483 285 8 https://doi.org/10.1145/3137597.3137600 https://doi.org/10.1145/3137597.3137600 ADD 12483 285 9 . . . 12483 286 1 23 23 CD 12483 286 2 Eugenio Eugenio NNP 12483 286 3 Tacchini Tacchini NNP 12483 286 4 et et FW 12483 286 5 al al NNP 12483 286 6 . . NNP 12483 286 7 , , , 12483 286 8 “ " `` 12483 286 9 Some some DT 12483 286 10 Like like VBP 12483 286 11 It -PRON- PRP 12483 286 12 Hoax Hoax NNP 12483 286 13 : : : 12483 286 14 Automated Automated NNP 12483 286 15 Fake Fake NNP 12483 286 16 News News NNP 12483 286 17 Detection Detection NNP 12483 286 18 in in IN 12483 286 19 Social Social NNP 12483 286 20 Networks Networks NNPS 12483 286 21 , , , 12483 286 22 ” " '' 12483 286 23 arXiv arXiv NNP 12483 286 24 preprint preprint NN 12483 286 25 arXiv:1704.07506 arXiv:1704.07506 NNP 12483 286 26 ( ( -LRB- 12483 286 27 2017 2017 CD 12483 286 28 ) ) -RRB- 12483 286 29 . . . 12483 287 1 24 24 CD 12483 287 2 Julio Julio NNP 12483 287 3 C.S. C.S. NNP 12483 288 1 Reis Reis NNP 12483 288 2 et et NNP 12483 288 3 al al NNP 12483 288 4 . . NNP 12483 288 5 , , , 12483 288 6 “ " `` 12483 288 7 Supervised supervise VBD 12483 288 8 Learning Learning NNP 12483 288 9 for for IN 12483 288 10 Fake Fake NNP 12483 288 11 News News NNP 12483 288 12 Detection detection NN 12483 288 13 , , , 12483 288 14 ” " '' 12483 288 15 IEEE IEEE NNP 12483 288 16 Intelligent Intelligent NNP 12483 288 17 Systems Systems NNPS 12483 288 18 34 34 CD 12483 288 19 , , , 12483 288 20 no no UH 12483 288 21 . . . 12483 289 1 2 2 CD 12483 289 2 ( ( -LRB- 12483 289 3 2019 2019 CD 12483 289 4 ) ) -RRB- 12483 289 5 : : : 12483 289 6 76–81 76–81 LS 12483 289 7 , , , 12483 289 8 https://doi.org10.1109 https://doi.org10.1109 NNP 12483 289 9 / / SYM 12483 289 10 MIS.2019.2899143 MIS.2019.2899143 NNP 12483 289 11 . . . 12483 290 1 25 25 CD 12483 290 2 Victoria Victoria NNP 12483 290 3 L. L. NNP 12483 290 4 Rubin Rubin NNP 12483 290 5 et et NNP 12483 290 6 al al NNP 12483 290 7 . . NNP 12483 290 8 , , , 12483 290 9 “ " `` 12483 290 10 Fake Fake NNP 12483 290 11 News News NNP 12483 290 12 or or CC 12483 290 13 Truth truth NN 12483 290 14 ? ? . 12483 291 1 Using use VBG 12483 291 2 Satirical satirical JJ 12483 291 3 Cues cue NNS 12483 291 4 to to TO 12483 291 5 Detect detect VB 12483 291 6 Potentially Potentially NNP 12483 291 7 Misleading Misleading NNP 12483 291 8 News News NNP 12483 291 9 , , , 12483 291 10 ” " '' 12483 291 11 ( ( -LRB- 12483 291 12 paper paper NN 12483 291 13 , , , 12483 291 14 Proceedings Proceedings NNPS 12483 291 15 of of IN 12483 291 16 the the DT 12483 291 17 Second Second NNP 12483 291 18 Workshop Workshop NNP 12483 291 19 on on IN 12483 291 20 Computational Computational NNP 12483 291 21 Approaches approach NNS 12483 291 22 to to TO 12483 291 23 Deception deception VB 12483 291 24 Detection detection NN 12483 291 25 , , , 12483 291 26 ( ( -LRB- 12483 291 27 2016 2016 CD 12483 291 28 ) ) -RRB- 12483 291 29 : : : 12483 291 30 7–17 7–17 NNP 12483 291 31 ) ) -RRB- 12483 291 32 ; ; : 12483 291 33 Francesco Francesco NNP 12483 291 34 Barbieri Barbieri NNP 12483 291 35 , , , 12483 291 36 Francesco Francesco NNP 12483 291 37 Ronzano Ronzano NNP 12483 291 38 , , , 12483 291 39 and and CC 12483 291 40 Horacio Horacio NNP 12483 291 41 Saggion Saggion NNP 12483 291 42 , , , 12483 291 43 “ " `` 12483 291 44 Is be VBZ 12483 291 45 This this DT 12483 291 46 Tweet Tweet NNP 12483 291 47 Satirical Satirical NNP 12483 291 48 ? ? . 12483 292 1 A a DT 12483 292 2 Computational Computational NNP 12483 292 3 Approach Approach NNP 12483 292 4 for for IN 12483 292 5 Satire Satire NNP 12483 292 6 Detection Detection NNP 12483 292 7 in in IN 12483 292 8 Spanish Spanish NNP 12483 292 9 , , , 12483 292 10 ” " '' 12483 292 11 Procesamiento Procesamiento NNP 12483 292 12 del del NNP 12483 292 13 Lenguaje Lenguaje NNP 12483 292 14 Natural Natural NNP 12483 292 15 , , , 12483 292 16 no no UH 12483 292 17 . . . 12483 293 1 55 55 CD 12483 293 2 ( ( -LRB- 12483 293 3 2015 2015 CD 12483 293 4 ) ) -RRB- 12483 293 5 : : : 12483 293 6 135 135 CD 12483 293 7 - - SYM 12483 293 8 42 42 CD 12483 293 9 ; ; : 12483 293 10 Soujanya Soujanya NNP 12483 293 11 Poria Poria NNP 12483 293 12 et et FW 12483 293 13 al al NNP 12483 293 14 . . NNP 12483 293 15 , , , 12483 293 16 “ " `` 12483 293 17 A a DT 12483 293 18 Deeper deep JJR 12483 293 19 Look look VB 12483 293 20 into into IN 12483 293 21 Sarcastic sarcastic JJ 12483 293 22 Tweets tweet NNS 12483 293 23 Using use VBG 12483 293 24 Deep Deep NNP 12483 293 25 Convolutional Convolutional NNP 12483 293 26 Neural Neural NNP 12483 293 27 Networks Networks NNPS 12483 293 28 , , , 12483 293 29 ” " '' 12483 293 30 arXiv arxiv JJ 12483 293 31 preprint preprint NN 12483 293 32 arXiv:1610.08815 arXiv:1610.08815 NNP 12483 293 33 ( ( -LRB- 12483 293 34 2016 2016 CD 12483 293 35 ) ) -RRB- 12483 293 36 . . . 12483 294 1 26 26 CD 12483 294 2 Lei Lei NNP 12483 294 3 Guo Guo NNP 12483 294 4 and and CC 12483 294 5 Chris Chris NNP 12483 294 6 Vargo Vargo NNP 12483 294 7 , , , 12483 294 8 “ " `` 12483 294 9 ’ ' '' 12483 294 10 Fake Fake NNP 12483 294 11 News News NNP 12483 294 12 ’ ' '' 12483 294 13 and and CC 12483 294 14 Emerging emerge VBG 12483 294 15 Online online JJ 12483 294 16 Media medium NNS 12483 294 17 Ecosystem Ecosystem NNP 12483 294 18 : : : 12483 294 19 An an DT 12483 294 20 Integrated Integrated NNP 12483 294 21 Intermedia Intermedia NNP 12483 294 22 Agenda Agenda NNP 12483 294 23 - - HYPH 12483 294 24 Setting Setting NNP 12483 294 25 Analysis Analysis NNP 12483 294 26 of of IN 12483 294 27 the the DT 12483 294 28 2016 2016 CD 12483 294 29 Us Us NNP 12483 294 30 Presidential Presidential NNP 12483 294 31 Election Election NNP 12483 294 32 , , , 12483 294 33 ” " '' 12483 294 34 Communication Communication NNP 12483 294 35 Research Research NNP 12483 294 36 47 47 CD 12483 294 37 , , , 12483 294 38 no no UH 12483 294 39 . . . 12483 295 1 2 2 CD 12483 295 2 ( ( -LRB- 12483 295 3 2020 2020 CD 12483 295 4 ) ) -RRB- 12483 295 5 : : : 12483 295 6 178–200 178–200 CD 12483 295 7 , , , 12483 295 8 https://doi.org/10.1177/0093650218777177 https://doi.org/10.1177/0093650218777177 ADD 12483 295 9 . . . 12483 296 1 27 27 CD 12483 296 2 Natali Natali NNP 12483 296 3 Ruchansky Ruchansky NNP 12483 296 4 , , , 12483 296 5 Sungyong Sungyong NNP 12483 296 6 Seo Seo NNP 12483 296 7 , , , 12483 296 8 and and CC 12483 296 9 Yan Yan NNP 12483 296 10 Liu Liu NNP 12483 296 11 , , , 12483 296 12 “ " `` 12483 296 13 CSI CSI NNP 12483 296 14 : : : 12483 296 15 A a DT 12483 296 16 Hybrid Hybrid NNP 12483 296 17 Deep Deep NNP 12483 296 18 Model model NN 12483 296 19 for for IN 12483 296 20 Fake Fake NNP 12483 296 21 News News NNP 12483 296 22 Detection detection NN 12483 296 23 , , , 12483 296 24 ” " '' 12483 296 25 Proceedings proceeding NNS 12483 296 26 of of IN 12483 296 27 the the DT 12483 296 28 2017 2017 CD 12483 296 29 ACM ACM NNP 12483 296 30 on on IN 12483 296 31 Conference Conference NNP 12483 296 32 on on IN 12483 296 33 Information Information NNP 12483 296 34 and and CC 12483 296 35 Knowledge Knowledge NNP 12483 296 36 Management Management NNP 12483 296 37 , , , 12483 296 38 ( ( -LRB- 12483 296 39 November November NNP 12483 296 40 2017 2017 CD 12483 296 41 ) ) -RRB- 12483 296 42 : : : 12483 296 43 797–806 797–806 CD 12483 296 44 , , , 12483 296 45 https://doi.org/10.1145/3132847.3132877 https://doi.org/10.1145/3132847.3132877 ADD 12483 296 46 . . . 12483 297 1 28 28 CD 12483 297 2 Yaqing Yaqing NNP 12483 297 3 Wang Wang NNP 12483 297 4 et et FW 12483 297 5 al al NNP 12483 297 6 . . NNP 12483 297 7 , , , 12483 297 8 “ " `` 12483 297 9 EANN eann NN 12483 297 10 : : : 12483 297 11 Event event NN 12483 297 12 Adversarial Adversarial NNP 12483 297 13 Neural Neural NNP 12483 297 14 Networks Networks NNPS 12483 297 15 for for IN 12483 297 16 Multi Multi NNP 12483 297 17 - - HYPH 12483 297 18 Modal Modal NNP 12483 297 19 Fake Fake NNP 12483 297 20 News News NNP 12483 297 21 Detection detection NN 12483 297 22 , , , 12483 297 23 ” " '' 12483 297 24 Proceedings proceeding NNS 12483 297 25 of of IN 12483 297 26 the the DT 12483 297 27 24th 24th JJ 12483 297 28 ACM ACM NNP 12483 297 29 SIGKDD SIGKDD NNP 12483 297 30 International International NNP 12483 297 31 Conference Conference NNP 12483 297 32 on on IN 12483 297 33 Knowledge Knowledge NNP 12483 297 34 Discovery Discovery NNP 12483 297 35 & & CC 12483 297 36 Data Data NNP 12483 297 37 Mining Mining NNP 12483 297 38 , , , 12483 297 39 ( ( -LRB- 12483 297 40 2018 2018 CD 12483 297 41 ) ) -RRB- 12483 297 42 : : : 12483 297 43 849–57 849–57 CD 12483 297 44 , , , 12483 297 45 https://doi.org/10.1145/3219819.3219903 https://doi.org/10.1145/3219819.3219903 ADD 12483 297 46 ; ; : 12483 297 47 James James NNP 12483 297 48 W. W. NNP 12483 297 49 Pennebaker Pennebaker NNP 12483 297 50 , , , 12483 297 51 Martha Martha NNP 12483 297 52 E. E. NNP 12483 297 53 Francis Francis NNP 12483 297 54 , , , 12483 297 55 and and CC 12483 297 56 Roger Roger NNP 12483 297 57 J. J. NNP 12483 297 58 Booth Booth NNP 12483 297 59 , , , 12483 297 60 “ " `` 12483 297 61 Linguistic linguistic JJ 12483 297 62 Inquiry Inquiry NNP 12483 297 63 and and CC 12483 297 64 Word Word NNP 12483 297 65 Count Count NNP 12483 297 66 : : : 12483 297 67 LIWC LIWC NNP 12483 297 68 2001 2001 CD 12483 297 69 ” " '' 12483 297 70 , , , 12483 297 71 Mahway Mahway NNP 12483 297 72 : : : 12483 297 73 Lawrence Lawrence NNP 12483 297 74 Erlbaum Erlbaum NNP 12483 297 75 Associates Associates NNP 12483 297 76 71 71 CD 12483 297 77 , , , 12483 297 78 no no UH 12483 297 79 . . . 12483 298 1 2001 2001 CD 12483 298 2 ( ( -LRB- 12483 298 3 2001 2001 CD 12483 298 4 ) ) -RRB- 12483 298 5 . . . 12483 299 1 29 29 CD 12483 299 2 “ " `` 12483 299 3 Facebook Facebook NNP 12483 299 4 , , , 12483 299 5 Twitter Twitter NNP 12483 299 6 May May NNP 12483 299 7 Face face VB 12483 299 8 More More JJR 12483 299 9 Scrutiny scrutiny NN 12483 299 10 in in IN 12483 299 11 2019 2019 CD 12483 299 12 to to TO 12483 299 13 Check check VB 12483 299 14 Fake Fake NNP 12483 299 15 News News NNP 12483 299 16 , , , 12483 299 17 Hate Hate NNP 12483 299 18 Speech Speech NNP 12483 299 19 , , , 12483 299 20 ” " '' 12483 299 21 accessed access VBD 12483 299 22 May May NNP 12483 299 23 17 17 CD 12483 299 24 , , , 12483 299 25 2020 2020 CD 12483 299 26 , , , 12483 299 27 available available JJ 12483 299 28 : : : 12483 299 29 https://www.huffingtonpost.in/entry/facebook-twitter-may-face- https://www.huffingtonpost.in/entry/facebook-twitter-may-face- ADD 12483 299 30 more more JJR 12483 299 31 - - HYPH 12483 299 32 scrutiny scrutiny NN 12483 299 33 - - HYPH 12483 299 34 in-2019-to in-2019-to NN 12483 299 35 - - HYPH 12483 299 36 check check VB 12483 299 37 - - HYPH 12483 299 38 fake fake JJ 12483 299 39 - - HYPH 12483 299 40 news news NN 12483 299 41 - - HYPH 12483 299 42 hate hate NN 12483 299 43 - - HYPH 12483 299 44 speech_in_5c29c589e4b05c88b701d72e speech_in_5c29c589e4b05c88b701d72e NN 12483 299 45 . . . 12483 300 1 30 30 CD 12483 300 2 Nitesh Nitesh NNP 12483 300 3 V. V. NNP 12483 300 4 Chawla Chawla NNP 12483 300 5 et et FW 12483 300 6 al al NNP 12483 300 7 . . NNP 12483 300 8 , , , 12483 300 9 “ " `` 12483 300 10 Smote Smote NNP 12483 300 11 : : : 12483 300 12 Synthetic Synthetic NNP 12483 300 13 Minority Minority NNP 12483 300 14 Over Over NNP 12483 300 15 - - HYPH 12483 300 16 Sampling Sampling NNP 12483 300 17 Technique Technique NNP 12483 300 18 , , , 12483 300 19 ” " '' 12483 300 20 Journal Journal NNP 12483 300 21 of of IN 12483 300 22 Artificial Artificial NNP 12483 300 23 Intelligence Intelligence NNP 12483 300 24 Research Research NNP 12483 300 25 16 16 CD 12483 300 26 , , , 12483 300 27 ( ( -LRB- 12483 300 28 2002 2002 CD 12483 300 29 ) ) -RRB- 12483 300 30 : : : 12483 300 31 321–57 321–57 CD 12483 300 32 , , , 12483 300 33 https://doi.org/10.1613/jair.953 https://doi.org/10.1613/jair.953 NNS 12483 300 34 . . . 12483 301 1 31 31 CD 12483 301 2 Uğur Uğur NNP 12483 301 3 Mertoğlu Mertoğlu NNP 12483 301 4 and and CC 12483 301 5 Burkay Burkay NNP 12483 301 6 Genç Genç NNP 12483 301 7 , , , 12483 301 8 “ " `` 12483 301 9 Lexicon Lexicon NNP 12483 301 10 Generation Generation NNP 12483 301 11 for for IN 12483 301 12 Detecting detect VBG 12483 301 13 Fake Fake NNP 12483 301 14 News News NNP 12483 301 15 , , , 12483 301 16 ” " '' 12483 301 17 arXiv arXiv NNP 12483 301 18 preprint preprint NN 12483 301 19 arXiv:2010.11089 arXiv:2010.11089 NNP 12483 301 20 ( ( -LRB- 12483 301 21 2020 2020 CD 12483 301 22 ) ) -RRB- 12483 301 23 . . . 12483 302 1 32 32 CD 12483 302 2 Burak Burak NNP 12483 302 3 Bezirci Bezirci NNP 12483 302 4 , , , 12483 302 5 and and CC 12483 302 6 Asım Asım NNP 12483 302 7 Egemen Egemen NNP 12483 302 8 Yilmaz Yilmaz NNP 12483 302 9 , , , 12483 302 10 “ " `` 12483 302 11 Metinlerin Metinlerin NNP 12483 302 12 Okunabilirliğinin Okunabilirliğinin NNP 12483 302 13 Ölçülmesi Ölçülmesi NNP 12483 302 14 Üzerine Üzerine NNP 12483 302 15 Bir Bir NNP 12483 302 16 Yazilim Yazilim NNP 12483 302 17 Kütüphanesi Kütüphanesi NNP 12483 302 18 Ve Ve NNP 12483 302 19 Türkçe Türkçe NNP 12483 302 20 Için Için NNP 12483 302 21 Yeni Yeni NNP 12483 302 22 Bir Bir NNP 12483 302 23 Okunabilirlik Okunabilirlik NNP 12483 302 24 Ölçütü Ölçütü NNP 12483 302 25 , , , 12483 302 26 ” " '' 12483 302 27 Dokuz Dokuz NNP 12483 302 28 Eylül Eylül NNP 12483 302 29 Üniversitesi Üniversitesi NNP 12483 302 30 https://doi.org/10.1371/journal.pone.0105937 https://doi.org/10.1371/journal.pone.0105937 NN 12483 302 31 https://ink.library.smu.edu.sg/sis_research/4630 https://ink.library.smu.edu.sg/sis_research/4630 ADD 12483 302 32 https://doi.org/10.1145/3137597.3137600 https://doi.org/10.1145/3137597.3137600 ADD 12483 302 33 https://doi.org10.1109 https://doi.org10.1109 NNP 12483 302 34 / / SYM 12483 302 35 MIS.2019.2899143 MIS.2019.2899143 NNP 12483 302 36 https://doi.org/10.1177%2F0093650218777177 https://doi.org/10.1177%2F0093650218777177 NNP 12483 302 37 https://doi.org/10.1145/3132847.3132877 https://doi.org/10.1145/3132847.3132877 ADD 12483 302 38 https://doi.org/10.1145/3219819.3219903 https://doi.org/10.1145/3219819.3219903 ADD 12483 302 39 https://www.huffingtonpost.in/entry/facebook-twitter-may-face-more-scrutiny-in-2019-to-check-fake-news-hate-speech_in_5c29c589e4b05c88b701d72e https://www.huffingtonpost.in/entry/facebook-twitter-may-face-more-scrutiny-in-2019-to-check-fake-news-hate-speech_in_5c29c589e4b05c88b701d72e NNP 12483 302 40 https://www.huffingtonpost.in/entry/facebook-twitter-may-face-more-scrutiny-in-2019-to-check-fake-news-hate-speech_in_5c29c589e4b05c88b701d72e https://www.huffingtonpost.in/entry/facebook-twitter-may-face-more-scrutiny-in-2019-to-check-fake-news-hate-speech_in_5c29c589e4b05c88b701d72e NNP 12483 302 41 https://doi.org/10.1613/jair.953 https://doi.org/10.1613/jair.953 VBZ 12483 302 42 INFORMATION INFORMATION VBN 12483 302 43 TECHNOLOGY technology NN 12483 302 44 AND and CC 12483 302 45 LIBRARIES library NNS 12483 302 46 DECEMBER DECEMBER NNP 12483 302 47 2020 2020 CD 12483 302 48 AUTOMATED AUTOMATED NNP 12483 302 49 FAKE FAKE NNP 12483 302 50 NEWS NEWS NNP 12483 302 51 DETECTION detection VB 12483 302 52 IN in IN 12483 302 53 THE the DT 12483 302 54 AGE AGE NNP 12483 302 55 OF of IN 12483 302 56 DIGITAL DIGITAL NNP 12483 302 57 LIBRARIES library NNS 12483 302 58 | | NNP 12483 302 59 MERTOĞLU MERTOĞLU NNS 12483 302 60 AND and CC 12483 302 61 GENÇ genç VBP 12483 302 62 19 19 CD 12483 302 63 Mühendislik Mühendislik NNP 12483 302 64 Fakültesi Fakültesi NNP 12483 302 65 Fen Fen NNP 12483 302 66 ve ve JJ 12483 302 67 Mühendislik Mühendislik NNP 12483 302 68 Dergisi Dergisi NNP 12483 302 69 12 12 CD 12483 302 70 , , , 12483 302 71 no no UH 12483 302 72 . . . 12483 303 1 3 3 CD 12483 303 2 ( ( -LRB- 12483 303 3 2010 2010 CD 12483 303 4 ) ) -RRB- 12483 303 5 : : : 12483 303 6 49–62 49–62 CD 12483 303 7 , , , 12483 303 8 https://dergipark.org.tr/en/pub/deumffmd/issue/40831/492667 https://dergipark.org.tr/en/pub/deumffmd/issue/40831/492667 ADD 12483 303 9 . . . 12483 304 1 33 33 CD 12483 304 2 Robert Robert NNP 12483 304 3 Gunning Gunning NNP 12483 304 4 , , , 12483 304 5 “ " `` 12483 304 6 The the DT 12483 304 7 technique technique NN 12483 304 8 of of IN 12483 304 9 clear clear JJ 12483 304 10 writing writing NN 12483 304 11 , , , 12483 304 12 ” " '' 12483 304 13 Revised Revised NNP 12483 304 14 Edition Edition NNP 12483 304 15 , , , 12483 304 16 New New NNP 12483 304 17 York York NNP 12483 304 18 : : : 12483 304 19 McGraw McGraw NNP 12483 304 20 Hill Hill NNP 12483 304 21 , , , 12483 304 22 1968 1968 CD 12483 304 23 . . . 12483 305 1 34 34 CD 12483 305 2 Ender Ender NNP 12483 305 3 Ateşman Ateşman NNP 12483 305 4 , , , 12483 305 5 “ " `` 12483 305 6 Türkçede Türkçede NNS 12483 305 7 Okunabilirliğin Okunabilirliğin NNP 12483 305 8 Ölçülmesi Ölçülmesi NNP 12483 305 9 , , , 12483 305 10 ” " '' 12483 305 11 Dil Dil NNP 12483 305 12 Dergisi Dergisi NNP 12483 305 13 58 58 CD 12483 305 14 , , , 12483 305 15 no no UH 12483 305 16 . . . 12483 306 1 71–74 71–74 CD 12483 306 2 ( ( -LRB- 12483 306 3 1997 1997 CD 12483 306 4 ) ) -RRB- 12483 306 5 . . . 12483 307 1 https://dergipark.org.tr/en/pub/deumffmd/issue/40831/492667 https://dergipark.org.tr/en/pub/deumffmd/issue/40831/492667 NNP 12483 307 2 ABSTRACT ABSTRACT NNP 12483 307 3 INTRODUCTION INTRODUCTION NNP 12483 307 4 LITERATURE LITERATURE NNP 12483 307 5 REVIEW REVIEW NNP 12483 307 6 METHODOLOGY METHODOLOGY NNP 12483 307 7 Framework Framework NNP 12483 307 8 Model Model NNP 12483 307 9 Dataset dataset NN 12483 307 10 Features Features NNP 12483 307 11 EXPERIMENTAL EXPERIMENTAL NNP 12483 307 12 RESULTS result NNS 12483 307 13 AND and CC 12483 307 14 DISCUSSION discussion NN 12483 307 15 CONCLUSION CONCLUSION NNP 12483 307 16 AND and CC 12483 307 17 FUTURE FUTURE NNP 12483 307 18 WORK work NN 12483 307 19 ENDNOTES endnote NNS