id sid tid token lemma pos pc289g58f72 1 1 graph graph NOUN pc289g58f72 1 2 data datum NOUN pc289g58f72 1 3 analysis analysis NOUN pc289g58f72 1 4 have have AUX pc289g58f72 1 5 recently recently ADV pc289g58f72 1 6 attained attain VERB pc289g58f72 1 7 much much ADJ pc289g58f72 1 8 interests interest NOUN pc289g58f72 1 9 from from ADP pc289g58f72 1 10 the the DET pc289g58f72 1 11 statistics statistic NOUN pc289g58f72 1 12 community community NOUN pc289g58f72 1 13 for for ADP pc289g58f72 1 14 their their PRON pc289g58f72 1 15 capabilities capability NOUN pc289g58f72 1 16 to to PART pc289g58f72 1 17 perform perform VERB pc289g58f72 1 18 inferential inferential ADJ pc289g58f72 1 19 tasks task NOUN pc289g58f72 1 20 on on ADP pc289g58f72 1 21 types type NOUN pc289g58f72 1 22 of of ADP pc289g58f72 1 23 data datum NOUN pc289g58f72 1 24 that that PRON pc289g58f72 1 25 are be AUX pc289g58f72 1 26 prevailing prevail VERB pc289g58f72 1 27 at at ADP pc289g58f72 1 28 every every DET pc289g58f72 1 29 corner corner NOUN pc289g58f72 1 30 of of ADP pc289g58f72 1 31 our our PRON pc289g58f72 1 32 daily daily ADJ pc289g58f72 1 33 life life NOUN pc289g58f72 1 34 . . PUNCT pc289g58f72 2 1 another another DET pc289g58f72 2 2 hot hot ADJ pc289g58f72 2 3 topic topic NOUN pc289g58f72 2 4 is be AUX pc289g58f72 2 5 neural neural ADJ pc289g58f72 2 6 network network NOUN pc289g58f72 2 7 which which PRON pc289g58f72 2 8 has have AUX pc289g58f72 2 9 been be AUX pc289g58f72 2 10 applied apply VERB pc289g58f72 2 11 on on ADP pc289g58f72 2 12 various various ADJ pc289g58f72 2 13 disciplines discipline NOUN pc289g58f72 2 14 , , PUNCT pc289g58f72 2 15 like like ADP pc289g58f72 2 16 partial partial ADJ pc289g58f72 2 17 deferential deferential ADJ pc289g58f72 2 18 equation(pde equation(pde NOUN pc289g58f72 2 19 ) ) PUNCT pc289g58f72 2 20 , , PUNCT pc289g58f72 2 21 for for SCONJ pc289g58f72 2 22 their their PRON pc289g58f72 2 23 capabilities capability NOUN pc289g58f72 2 24 to to PART pc289g58f72 2 25 deal deal VERB pc289g58f72 2 26 with with ADP pc289g58f72 2 27 high high ADJ pc289g58f72 2 28 dimensional dimensional ADJ pc289g58f72 2 29 problems problem NOUN pc289g58f72 2 30 that that SCONJ pc289g58f72 2 31 traditional traditional ADJ pc289g58f72 2 32 finite finite NOUN pc289g58f72 2 33 elements element NOUN pc289g58f72 2 34 or or CCONJ pc289g58f72 2 35 finite finite ADJ pc289g58f72 2 36 difference difference NOUN pc289g58f72 2 37 methods method NOUN pc289g58f72 2 38 often often ADV pc289g58f72 2 39 fail fail VERB pc289g58f72 2 40 to to PART pc289g58f72 2 41 solve solve VERB pc289g58f72 2 42 . . PUNCT pc289g58f72 3 1 my my PRON pc289g58f72 3 2 dissertation dissertation NOUN pc289g58f72 3 3 develops develop VERB pc289g58f72 3 4 several several ADJ pc289g58f72 3 5 methods method NOUN pc289g58f72 3 6 that that PRON pc289g58f72 3 7 focus focus VERB pc289g58f72 3 8 on on ADP pc289g58f72 3 9 various various ADJ pc289g58f72 3 10 aspects aspect NOUN pc289g58f72 3 11 across across ADP pc289g58f72 3 12 these these DET pc289g58f72 3 13 two two NUM pc289g58f72 3 14 disciplines discipline NOUN pc289g58f72 3 15 . . PUNCT pc289g58f72 4 1 in in ADP pc289g58f72 4 2 the the DET pc289g58f72 4 3 graph graph NOUN pc289g58f72 4 4 - - PUNCT pc289g58f72 4 5 related relate VERB pc289g58f72 4 6 realm realm NOUN pc289g58f72 4 7 , , PUNCT pc289g58f72 4 8 we we PRON pc289g58f72 4 9 firstly firstly ADV pc289g58f72 4 10 developed develop VERB pc289g58f72 4 11 efficient efficient ADJ pc289g58f72 4 12 numerical numerical ADJ pc289g58f72 4 13 routines routine NOUN pc289g58f72 4 14 for for ADP pc289g58f72 4 15 unsupervised unsupervised ADJ pc289g58f72 4 16 clustering clustering NOUN pc289g58f72 4 17 problem problem NOUN pc289g58f72 4 18 and and CCONJ pc289g58f72 4 19 further far ADV pc289g58f72 4 20 proved prove VERB pc289g58f72 4 21 theoretical theoretical ADJ pc289g58f72 4 22 guarantees guarantee NOUN pc289g58f72 4 23 on on ADP pc289g58f72 4 24 the the DET pc289g58f72 4 25 clustering clustering NOUN pc289g58f72 4 26 error error NOUN pc289g58f72 4 27 corresponding correspond VERB pc289g58f72 4 28 to to ADP pc289g58f72 4 29 the the DET pc289g58f72 4 30 numerical numerical ADJ pc289g58f72 4 31 routines routine NOUN pc289g58f72 4 32 . . PUNCT pc289g58f72 5 1 for for ADP pc289g58f72 5 2 graph graph NOUN pc289g58f72 5 3 neural neural ADJ pc289g58f72 5 4 network network NOUN pc289g58f72 5 5 , , PUNCT pc289g58f72 5 6 we we PRON pc289g58f72 5 7 modified modify VERB pc289g58f72 5 8 the the DET pc289g58f72 5 9 traditional traditional ADJ pc289g58f72 5 10 training training NOUN pc289g58f72 5 11 process process NOUN pc289g58f72 5 12 of of ADP pc289g58f72 5 13 graph graph NOUN pc289g58f72 5 14 neural neural ADJ pc289g58f72 5 15 network network NOUN pc289g58f72 5 16 by by ADP pc289g58f72 5 17 re re NOUN pc289g58f72 5 18 - - VERB pc289g58f72 5 19 sampling sample VERB pc289g58f72 5 20 the the DET pc289g58f72 5 21 graph graph NOUN pc289g58f72 5 22 structure structure NOUN pc289g58f72 5 23 . . PUNCT pc289g58f72 6 1 to to ADP pc289g58f72 6 2 the the DET pc289g58f72 6 3 best good ADJ pc289g58f72 6 4 of of ADP pc289g58f72 6 5 our our PRON pc289g58f72 6 6 knowledge knowledge NOUN pc289g58f72 6 7 , , PUNCT pc289g58f72 6 8 we we PRON pc289g58f72 6 9 are be AUX pc289g58f72 6 10 the the DET pc289g58f72 6 11 first first ADJ pc289g58f72 6 12 to to PART pc289g58f72 6 13 introduce introduce VERB pc289g58f72 6 14 the the DET pc289g58f72 6 15 randomness randomness NOUN pc289g58f72 6 16 into into ADP pc289g58f72 6 17 the the DET pc289g58f72 6 18 training training NOUN pc289g58f72 6 19 phase phase NOUN pc289g58f72 6 20 of of ADP pc289g58f72 6 21 graph graph NOUN pc289g58f72 6 22 neural neural ADJ pc289g58f72 6 23 network network NOUN pc289g58f72 6 24 . . PUNCT pc289g58f72 7 1 on on ADP pc289g58f72 7 2 the the DET pc289g58f72 7 3 pde pde ADJ pc289g58f72 7 4 domain domain NOUN pc289g58f72 7 5 , , PUNCT pc289g58f72 7 6 we we PRON pc289g58f72 7 7 proposed propose VERB pc289g58f72 7 8 structure structure NOUN pc289g58f72 7 9 - - PUNCT pc289g58f72 7 10 preserving preserve VERB pc289g58f72 7 11 neural neural ADJ pc289g58f72 7 12 - - PUNCT pc289g58f72 7 13 network network NOUN pc289g58f72 7 14 - - PUNCT pc289g58f72 7 15 based base VERB pc289g58f72 7 16 numerical numerical ADJ pc289g58f72 7 17 schemes scheme NOUN pc289g58f72 7 18 to to PART pc289g58f72 7 19 solve solve VERB pc289g58f72 7 20 both both DET pc289g58f72 7 21 l2 l2 NOUN pc289g58f72 7 22 - - PUNCT pc289g58f72 7 23 gradient gradient NOUN pc289g58f72 7 24 flows flow NOUN pc289g58f72 7 25 and and CCONJ pc289g58f72 7 26 generalized generalized ADJ pc289g58f72 7 27 diffusions diffusion NOUN pc289g58f72 7 28 . . PUNCT pc289g58f72 8 1 by by ADP pc289g58f72 8 2 using use VERB pc289g58f72 8 3 neural neural ADJ pc289g58f72 8 4 networks network NOUN pc289g58f72 8 5 as as ADP pc289g58f72 8 6 tools tool NOUN pc289g58f72 8 7 for for ADP pc289g58f72 8 8 spatial spatial ADJ pc289g58f72 8 9 discretization discretization NOUN pc289g58f72 8 10 , , PUNCT pc289g58f72 8 11 we we PRON pc289g58f72 8 12 introduce introduce VERB pc289g58f72 8 13 a a DET pc289g58f72 8 14 structure structure NOUN pc289g58f72 8 15 - - PUNCT pc289g58f72 8 16 preserving preserve VERB pc289g58f72 8 17 eulerian eulerian ADJ pc289g58f72 8 18 algorithm algorithm NOUN pc289g58f72 8 19 to to PART pc289g58f72 8 20 solve solve VERB pc289g58f72 8 21 l2 l2 NOUN pc289g58f72 8 22 - - PUNCT pc289g58f72 8 23 gradient gradient NOUN pc289g58f72 8 24 flows flow NOUN pc289g58f72 8 25 and and CCONJ pc289g58f72 8 26 a a DET pc289g58f72 8 27 structure structure NOUN pc289g58f72 8 28 - - PUNCT pc289g58f72 8 29 preserving preserve VERB pc289g58f72 8 30 lagrangian lagrangian NOUN pc289g58f72 8 31 algorithm algorithm PROPN pc289g58f72 8 32 to to PART pc289g58f72 8 33 solve solve VERB pc289g58f72 8 34 generalized generalized ADJ pc289g58f72 8 35 diffusions diffusion NOUN pc289g58f72 8 36 . . PUNCT