bigram

This is a table of type bigram and their frequencies. Use it to search & browse the list to learn more about your study carrel.

bigram	frequency
social media	890
machine learning	212
public health	170
altmetric data	152
twitter data	132
twitter users	126
social networks	110
big data	107
research topics	104
united states	97
social distancing	85
brand equity	81
sentiment analysis	81
hate speech	79
social network	75
marketing assets	59
media data	58
using twitter	57
data collection	57
data mining	56
media coverage	56
doc id	54
natural language	54
cord uid	54
fake news	53
network analysis	52
media platforms	51
local tv	49
data analysis	49
hot research	47
information diffusion	45
mental health	45
higher levels	44
total number	43
event detection	43
conspiracy narratives	42
topic modeling	41
language processing	41
data sources	40
news frames	40
table shows	39
tv viewership	38
sentiment classification	36
twitter activity	36
million tweets	35
financial performance	35
negative tweets	34
game attendance	34
time period	34
news media	34
personal negative	34
online social	33
deep learning	33
new york	32
verified users	32
abusive replies	32
content analysis	31
twitter api	31
using social	31
time series	30
twitter followers	30
world health	30
see table	30
political orientation	29
data presence	29
health organization	29
disease outbreaks	28
twitter dataset	28
stakeholder value	28
novel coronavirus	28
panic buying	28
case study	28
tweets dataset	27
tweets containing	27
made available	27
author funder	27
granted medrxiv	27
right wing	27
twitter mentions	27
nan doi	27
information operations	27
political parties	26
international license	26
health crisis	26
subject fields	26
media use	25
text mining	25
peer review	25
validated network	25
related tweets	25
media literacy	25
copyright holder	25
statistically significant	25
social learning	25
semantic network	24
boris johnson	24
home schooling	24
personal tweets	24
social big	24
streaming api	24
media users	24
conspiracy theories	24
twitter accounts	24
abusive language	23
personal non	23
accuracy assessment	23
ticket sales	23
coronavirus disease	23
team financial	23
news tweet	23
confirmed cases	22
school closures	22
original tweets	22
use social	22
diffusion patterns	22
depression detection	22
artificial intelligence	22
infectious disease	22
news coverage	22
data set	22
twitter posts	22
uk mps	22
publicly available	22
risk perception	21
social networking	21
communication ecosystem	21
retweet network	21
among others	21
gender differences	21
tweets using	21
new media	20
law enforcement	20
nan sha	20
latent dirichlet	20
future research	20
topic modelling	20
league baseball	20
domestic abuse	20
public opinion	20
el mundo	20
sir model	20
international conference	20
major league	20
hashtag groups	20
depressed users	20
version posted	20
label propagation	20
per day	20
tweets posted	20
word embedding	19
deaths per	19
seasonal ma	19
english tweets	19
real time	19
location information	19
tweet ids	19
news articles	19
may also	19
twitter use	19
two different	18
donald trump	18
research question	18
regression models	18
dirichlet allocation	18
historical sharing	18
digital platforms	18
stress symptoms	18
crisis communication	18
state sponsored	18
abuse levels	18
measles outbreak	18
smt immunity	18
least one	18
immunity link	18
privacy concerns	18
previous studies	18
community detection	18
large number	17
health care	17
across different	17
twitter sentiment	17
word embeddings	17
twitter messages	17
word co	17
president trump	17
time periods	17
methods used	17
dependent variable	17
twitter account	17
high level	17
learning algorithms	17
frames used	16
data processing	16
digital media	16
general covid	16
disaster management	16
moral panics	16
twitter user	16
personal tweet	16
words related	16
bipartite network	16
may th	16
personal versus	16
false information	16
news tweets	16
tweets related	16
even though	16
information ecology	16
supplementary material	16
commonly used	16
epidemic intelligence	16
high levels	16
psychophysical numbing	16
moral panic	16
neural networks	16
semantic networks	16
open access	15
public concern	15
swine flu	15
results show	15
text analysis	15
early stages	15
data source	15
seeded lda	15
topic models	15
literature review	15
collect data	15
two weeks	15
conspiracy theory	15
information cascades	15
hurricane sandy	15
political party	15
different types	15
unified theory	15
medrxiv preprint	15
widely used	15
abuse toward	15
risk communication	15
classification method	15
different stages	15
fechner law	15
support vector	15
infectious diseases	15
tweets per	14
digital literacy	14
information sharing	14
bot detection	14
see appendix	14
quote tweets	14
prime minister	14
lda model	14
covid information	14
data hunger	14
news frame	14
user behaviour	14
early warning	14
political frame	14
subject field	14
recent years	14
season marketing	14
mining techniques	14
political polarization	14
english language	14
resilient communication	14
language tweets	14
power law	14
data across	14
retweet prediction	14
collected tweets	14
action words	14
previous research	14
versus news	14
posted may	14
via twitter	13
first step	13
reported likelihood	13
research questions	13
tweets collected	13
based approach	13
lower levels	13
hashtag group	13
corexq algorithm	13
original tweet	13
negative versus	13
health information	13
topic level	13
altmetric events	13
versus non	13
trend analysis	13
medical education	13
general public	13
opinion mining	13
disease surveillance	13
th inclusive	13
learning model	13
health events	13
learning methods	13
data coverage	13
mass media	13
raw data	13
maintenance stage	13
word cloud	13
community structure	13
virtue signalling	13
supplementary information	13
online abuse	13
potential reach	13
table gives	13
hunter tier	13
different methods	13
wos publications	13
negative tweet	13
seasonal marketing	13
twitter stream	13
large scale	13
mendeley readers	13
twitter content	13
italia viva	13
social science	13
sponsored media	12
data sets	12
refuting tweets	12
jaccard coefficient	12
april th	12
many people	12
two categories	12
future work	12
sample size	12
health frame	12
two types	12
political news	12
false material	12
highest number	12
policy document	12
sentiment score	12
social cybersecurity	12
via social	12
tweet text	12
media messages	12
detection using	12
see section	12
learning systems	12
three different	12
news update	12
key words	12
social web	12
non reputable	12
false news	12
neural network	12
negative sentiment	12
also used	12
million users	12
twitter provides	12
natural disasters	12
firm value	12
networking sites	12
replies received	12
related terms	12
geocov tweets	12
facebook mentions	12
social data	12
media activities	12
political discourse	12
modal features	12
shifting individuals	12
steel blue	12
online hate	12
linguistic analysis	12
previous work	12
public discourse	12
toilet paper	12
climate change	12
across subject	11
recovery period	11
learning models	11
early days	11
health research	11
collected data	11
frequently used	11
tested positive	11
less likely	11
mendeley readership	11
fuzzy accuracy	11
march th	11
update frame	11
every day	11
reply tweets	11
crime prevention	11
news mentions	11
vaccine frame	11
one another	11
presidential election	11
per number	11
words associated	11
exploratory analysis	11
mental illness	11
public perception	11
also observe	11
expanded analysis	11
different political	11
analysis methods	11
political stories	11
two main	11
logistic regression	11
york times	11
public healthcare	11
global pandemic	11
topic model	11
public sentiment	11
mention network	11
team financials	11
abc news	11
united kingdom	11
will also	11
common words	11
training data	11
automated accounts	11
mainstream media	11
tw i	11
based method	11
news classification	11
news sources	11
unverified users	11
social bots	11
official accounts	11
untrue material	10
news count	10
personality traits	10
different countries	10
early detection	10
different social	10
related words	10
marketing asset	10
based approaches	10
disease outbreak	10
use twitter	10
red line	10
research areas	10
pattern recognition	10
remote learning	10
following tweet	10
us presidential	10
may lead	10
phrase i	10
blogs citations	10
depression symptoms	10
worth noting	10
false positives	10
large networks	10
diffusion dynamics	10
public response	10
descriptive statistics	10
information retrieval	10
standard deviation	10
season ma	10
hate generation	10
may help	10
research topic	10
open source	10
many tweets	10
word clouds	10
main categories	10
allows us	10
th april	10
april st	10
confusion matrix	10
statistically significantly	10
also show	10
emerging infectious	10
analysis using	10
time frame	10
vast majority	10
experimental results	10
daily number	10
health issues	10
media platform	10
abuse level	10
table presents	10
health professional	10
speech detection	10
classification system	10
vector machine	10
will use	9
media twitter	9
analysis techniques	9
disease control	9
policy makers	9
profanity words	9
saudi arabia	9
clustering algorithm	9
conspiracy beliefs	9
named entities	9
analysis results	9
political events	9
higher likelihood	9
third party	9
presence across	9
four studies	9
time window	9
table table	9
wing parties	9
crisis period	9
agenda setting	9
seed words	9
labour politicians	9
text data	9
around covid	9
directed validated	9
exponential growth	9
two periods	9
also found	9
government agencies	9
death rate	9
authors also	9
exogenous signals	9
positively related	9
thematic analysis	9
retweet cascades	9
protective equipment	9
active users	9
higher extraversion	9
twitter rest	9
news timeline	9
topic detection	9
one study	9
example tweets	9
negative classification	9
news detection	9
verified user	9
new information	9
rest api	9
popular hashtags	9
crisis management	9
european union	9
lockdown period	9
specific altmetric	9
lower agreeableness	9
better understand	9
uk news	9
first two	9
trending hashtags	9
systematic review	9
also known	9
detecting influenza	9
surveillance system	9
twitter using	9
gives us	9
abuse sent	9
different altmetric	9
shared untrue	9
uk politicians	9
cnn model	9
usa states	9
online news	9
tweets concerned	9
text classification	9
unique users	9
cov tweets	9
last two	9
pew research	9
based methods	9
features representing	9
naive bayes	9
different time	9
current work	9
health crises	9
research outputs	9
document citations	9
data points	9
disaster risk	9
centrality metrics	9
virus spread	9
information warfare	9
stay home	9
scientific publications	9
multiple regression	9
outbreak detection	9
quantitative analysis	9
future studies	8
verified twitter	8
evacuation orders	8
sentiment polarity	8
disaster response	8
clinical scenarios	8
phq value	8
terms used	8
keystone species	8
learning framework	8
greater likelihood	8
risk reduction	8
zero values	8
available data	8
jeremy corbyn	8
related hashtags	8
disinformation stories	8
omitted variables	8
sharing disinformation	8
authors declare	8
diversionary tweets	8
activity history	8
file system	8
mapreduce paradigm	8
news source	8
learning approach	8
allows users	8
last years	8
one million	8
will change	8
twitter event	8
also see	8
political authority	8
spinal manipulation	8
mlb teams	8
generated content	8
tweets mentioning	8
least squares	8
covid crisis	8
skewed distribution	8
hateful tweets	8
main topics	8
disinformation items	8
many times	8
review comments	8
relatively small	8
modalities attribute	8
twitter streaming	8
online engagement	8
twitter date	8
healthcare workers	8
study period	8
prominent news	8
relevant information	8
non seasonal	8
public concerns	8
initial stage	8
influenza epidemics	8
web data	8
word count	8
related stress	8
information regarding	8
media manipulation	8
stop words	8
relatively high	8
information network	8
per one	8
pandemic update	8
sky blue	8
allow users	8
meaning accretion	8
news headlines	8
word frequency	8
different data	8
information technology	8
across social	8
similar research	8
results also	8
based classification	8
user experiences	8
user engagement	8
twitter communications	8
real world	8
quote tweet	8
marketing analytics	8
textual data	8
decision support	8
echo chambers	8
social sciences	8
regression analysis	8
per post	8
pilot work	8
positive relationship	8
different sources	8
tweets promoting	8
closely related	8
occurrence networks	8
significantly predicted	8
mlb brand	8
data collected	8
epidemic outbreaks	8
profanity list	8
sampling period	8
million people	8
term frequency	8
journal clubs	8
negative score	8
average sentiment	8
visual analytics	8
relatively low	8
behavior analysis	8
training dataset	8
news story	8
spatial sensors	8
media sites	8
user tweets	8
three disinformation	8
significantly higher	8
citation data	8
social world	8
endoscopic evaluation	8
data analytics	8
news reports	8
qualitative analysis	8
media user	8
personal protective	8
moderating effect	8
police forces	8
reported symptoms	8
typhoon hagibis	7
left wing	7
one hundred	7
language use	7
next section	7
south korea	7
targeted analysis	7
tweet dataset	7
topics related	7
regression model	7
using different	7
respiratory syndrome	7
programming interface	7
published maps	7
inverse document	7
related corpora	7
personal information	7
professional sports	7
sporting events	7
mueller coverage	7
user will	7
social change	7
related information	7
public coronavirus	7
hybrid deep	7
arabic tweets	7
blue community	7
verbal abuse	7
axis represents	7
tweets may	7
bipartite directed	7
existing beliefs	7
worth mentioning	7
twitter network	7
general locations	7
table i	7
policy documents	7
pattern pool	7
topics generated	7
health concerns	7
health surveillance	7
common symptoms	7
crucial role	7
tweet contains	7
negative matrix	7
better understanding	7
detection algorithms	7
digital communications	7
mueller investigation	7
twitter vocabulary	7
null model	7
wing community	7
save lives	7
matrix factorization	7
note springer	7
sentiment monitoring	7
degree sequence	7
increasing number	7
small news	7
positive users	7
recurrent neural	7
every user	7
top ten	7
media outlets	7
grained linguistic	7
dependency parsing	7
science research	7
twitter discussion	7
also provides	7
acute respiratory	7
social life	7
single tweet	7
application programming	7
data technologies	7
tweets refuting	7
sharing information	7
classification problem	7
section iv	7
empirical studies	7
monitoring public	7
tweet data	7
springer nature	7
jurisdictional claims	7
per minute	7
financial support	7
media systems	7
second step	7
media mining	7
human perception	7
discussion around	7
retweet cascade	7
remains neutral	7
control variables	7
chinese government	7
frequent words	7
flu trends	7
document frequency	7
word vectors	7
lockdown measures	7
eigenvector centrality	7
different communities	7
graph processing	7
severe acute	7
online platforms	7
malicious content	7
give us	7
missile test	7
disease ontology	7
political sub	7
social sensors	7
positive rate	7
hybrid model	7
news item	7
probability distribution	7
media discourse	7
institutional affiliations	7
online political	7
motivation theory	7
first three	7
section describes	7
cwts classification	7
next step	7
people use	7
table provides	7
many twitter	7
using natural	7
attention received	7
individual tweets	7
reported cases	7
many countries	7
world events	7
recent research	7
decision tree	7
important information	7
statistical analysis	7
results obtained	7
keyword analysis	7
named entity	7
will discuss	7
top words	7
based social	7
death nls	7
key value	7
facebook use	7
provide information	7
media sources	7
team brand	7
high abuse	7
surveillance systems	7
false political	7
spatial uncertainty	7
first time	7
research center	7
collected twitter	7
infected cases	7
johns hopkins	7
coronavirus twitter	7
fake accounts	7
relationships among	7
general election	7
detailed information	7
data using	7
creative commons	7
classification using	7
true information	7
series analysis	7
significant predictor	7
several times	7
hashtags related	7
step classification	7
relevant tweets	7
input data	7
analysis tasks	7
top hashtags	7
word vec	7
attention mechanism	7
also included	7
spanish newspapers	7
tagged tweets	7
epidemic outbreak	7
like twitter	7
information spread	7
non trivial	7
may increase	7
search engine	7
media activity	7
medical conferences	7
based brand	7
health problems	7
recent studies	7
real network	7
distancing rules	7
search terms	7
allowed us	7
four main	7
protection motivation	7
nature remains	7
journal club	7
terms related	7
active twitter	7
media posts	7
discussion topics	6
data shows	6
different phases	6
weakly supervised	6
information provided	6
will return	6
global graph	6
existing studies	6
depressive symptoms	6
semantic features	6
influenza epidemic	6
fuzzy set	6
control group	6
mistakenly released	6
decision making	6
weaker relationship	6
collective action	6
analyzing twitter	6
ebola outbreak	6
data model	6
median time	6
use cases	6
network dynamics	6
historical tweets	6
best performing	6
strongly subjective	6
news stories	6
social issues	6
interestingly enough	6
weakly subjective	6
urgent endoscopy	6
mlb attendance	6
harmful media	6
detection methods	6
trending topics	6
web page	6
early february	6
learning techniques	6
specific hashtags	6
conservative party	6
public information	6
context extraction	6
concern timeline	6
opening remarks	6
bursty phrases	6
york city	6
user data	6
german speaking	6
identify hot	6
proposed algorithm	6
current findings	6
chinese social	6
class imbalance	6
social isolation	6
tweets based	6
british mps	6
epidemiological modeling	6
google news	6
i know	6
search api	6
health informatics	6
extraction patterns	6
geographic regions	6
high dimensional	6
ebola virus	6
semantic web	6
prior exposure	6
research studies	6
six main	6
facebook users	6
irrelevant tweets	6
data stream	6
baseball attendance	6
daily death	6
coronavirus topic	6
posted june	6
story attributed	6
social impact	6
word pair	6
important issue	6
previous subsection	6
paper presents	6
deliberate sharing	6
gated recurrent	6
configuration model	6
big social	6
section provides	6
three symptoms	6
test surge	6
large amount	6
tab delimited	6
temporal dynamics	6
story online	6
news items	6
health organizations	6
spatial distribution	6
intensive care	6
effect sizes	6
john hopkins	6
educated people	6
randomly sampled	6
secondary data	6
dark orange	6
news outlets	6
user posts	6
hate diffusion	6
media infodemic	6
semantic role	6
ground truth	6
political situation	6
make sense	6
instagram users	6
lower conscientiousness	6
also consider	6
thematic maps	6
lower age	6
dictionary learning	6
missing data	6
ecology framework	6
diffusion models	6
initial tweet	6
media briefing	6
predictive models	6
knowledge discovery	6
high precision	6
twitter journal	6
previous works	6
crime reduction	6
times higher	6
positive tweets	6
undirected version	6
global crisis	6
perceptual screen	6
positive cases	6
given time	6
relevant data	6
zika virus	6
blue subcommunity	6
word pairs	6
negative words	6
role labeling	6
step sentiment	6
rated likelihood	6
unsupervised machine	6
traditional news	6
study shows	6
public policies	6
dimensionality reduction	6
towards uk	6
foreign residents	6
data used	6
folk devils	6
citation counts	6
media content	6
lda models	6
see methods	6
information may	6
hopkins university	6
manually reviewed	6
scale covid	6
vaccine sentiment	6
social processes	6
mutual information	6
chinese propaganda	6
take care	6
potentially harmful	6
exogenous attention	6
classify tweets	6
eid outbreak	6
two newspapers	6
detecting depression	6
user profile	6
white house	6
coronavirus resource	6
emotional response	6
health emergencies	6
toward mps	6
timely manner	6
toward uk	6
user id	6
geo version	6
intensive attention	6
situational awareness	6
disease activity	6
computational intelligence	6
network sites	6
flu pandemic	6
meeting hashtag	6
series data	6
liwc dictionary	6
crisis events	6
first week	6
tweets sent	6
national security	6
th march	6
regression coefficients	6
information sources	6
disaster resilience	6
clinical settings	6
data show	6
subcommunity contains	6
linear model	6
negative binomial	6
detection model	6
twitter corpora	6
psychophysical laws	6
topic coherence	6
based null	6
related messages	6
based model	6
better performance	6
general crime	6
annotated datasets	6
total correlation	6
results showed	6
online information	6
specific data	6
main text	6
pandemic date	6
related work	6
linguistic features	6
disaster relief	6
table ii	6
offensive output	6
recent work	6
study using	6
word set	6
south america	6
slightly different	6
retweet dynamics	6
stress words	6
relevant nodes	6
epidemics using	6
health agencies	6
another study	6
help us	6
location spoofing	6
general elections	6
american society	6
feature extraction	6
main subject	6
dominic cummings	6
information shared	6
negative emotions	6
platforms like	6
entity recognition	6
will help	6
current flow	6
key topics	6
predictive analytics	6
will get	6
target audience	6
removed cases	6
media communication	6
tweets contained	6
different hashtags	6
covid tweets	6
untrue political	6
equity model	6
digital traces	6
opposition parties	5
used twitter	5
test dataset	5
digital age	5
disaster communication	5
matt hancock	5
human annotated	5
one possible	5
communication strategies	5
predictor variable	5
world population	5
annual meeting	5
deliberate historical	5
network data	5
political disinformation	5
two layers	5
near real	5
previous method	5
crime types	5
personal reports	5
analyzing tweets	5
mainly composed	5
observed pattern	5
level features	5
research articles	5
one hand	5
com jacklopresti	5
rss feeds	5
observational study	5
different topics	5
detecting public	5
later found	5
random forest	5
february th	5
language toolkit	5
detection method	5
also important	5
tweet received	5
will need	5
three periods	5
people also	5
news peaks	5
news agencies	5
tweets received	5
information fusion	5
test whether	5
reputable sources	5
also called	5
online debate	5
results presented	5
user base	5
data science	5
among users	5
tweets without	5
health facilities	5
critical equipment	5
unclear whether	5
current covid	5
labeling method	5
republican parties	5
staying home	5
presidential campaign	5
multiple sources	5
brexit coverage	5
death cases	5
young adults	5
retweeting times	5
community resilience	5
python package	5
public safety	5
feature representation	5
people reporting	5
multiple studies	5
two attributes	5
demographic characteristics	5
late may	5
disease detection	5
distribution patterns	5
timeline trend	5
global public	5
south alabama	5
recently proposed	5
german language	5
sentiment scores	5
sports marketing	5
considerable amount	5
phq level	5
three times	5
using two	5
subjective clues	5
promoting tweets	5
online communication	5
organic diffusion	5
data privacy	5
average number	5
hateful behavior	5
news sites	5
crisis stage	5
covid related	5
across research	5
decision makers	5
users tend	5
first research	5
basilisk algorithm	5
profanity word	5
political agenda	5
user mentions	5
user interactions	5
papers published	5
model using	5
around million	5
tweeting practices	5
media using	5
physical magnitude	5
twitter datasets	5
prediction task	5
specific words	5
different frames	5
existing literature	5
political memes	5
manual grouping	5
network structure	5
one week	5
news feed	5
terrorist attacks	5
concerns regarding	5
sentiment timeline	5
distributed storage	5
events related	5
topic embedding	5
coherence measure	5
correlation explanation	5
web sites	5
mechanical turk	5
specific features	5
biggest communities	5
compromised accounts	5
boolean terms	5
twitter based	5
public relations	5
complex networks	5
hate detection	5
articles published	5
sparse matrix	5
yemen cholera	5
sentiment classifier	5
using machine	5
mental disorders	5
across multiple	5
study found	5
large amounts	5
political communication	5
search results	5
stress level	5
knowledge dissemination	5
altmetric event	5
trending topic	5
latent features	5
two frames	5
intensity values	5
non verified	5
cognitive systems	5
social relations	5
whether tweets	5
clustering algorithms	5
information processing	5
liberal democrats	5
trauma surgery	5
based models	5
online media	5
search engines	5
shared material	5
current study	5
web mining	5
vector space	5
stronger relationship	5
description see	5
public responses	5
user timeline	5
epidemiological data	5
publication level	5
clinical studies	5
words across	5
public attention	5
topological properties	5
specific tweets	5
using data	5
feature vector	5
retweeter prediction	5
processing techniques	5
public opinions	5
mining social	5
better results	5
moving average	5
scientific research	5
social consequences	5
spectral clustering	5
studies based	5
related news	5
share false	5
news content	5
daily life	5
like characteristics	5
information available	5
coronavirus outbreak	5
team performance	5
quantitative performance	5
possible explanation	5
data quality	5
temporal trends	5
crimson hexagon	5
high number	5
receiving abuse	5
often used	5
tom buchanan	5
health communication	5
least symptom	5
national daily	5
higher level	5
speaking twitter	5
crime data	5
potential retweeters	5
revenue sharing	5
many different	5
specific topics	5
mining twitter	5
hateful content	5
semantic space	5
participant demographics	5
research interests	5
table lists	5
validated projection	5
media space	5
health sciences	5
studies related	5
data volume	5
help understand	5
popular topics	5
wikipedia citations	5
scale social	5
risk information	5
place managers	5
reddit mentions	5
world series	5
several limitations	5
increasing use	5
hateful contents	5
twitter analysis	5
one day	5
new method	5
formal outbreak	5
one user	5
gender difference	5
disaster events	5
time i	5
relevant terms	5
fast unfolding	5
march st	5
testing positive	5
th may	5
center left	5
increased cases	5
candidate word	5
small number	5
identifying hot	5
best results	5
viral marketing	5
korea missile	5
show different	5
dynamic setting	5
trained word	5
effect size	5
diffusion process	5
twitter information	5
identified hot	5
information flows	5
digitally mediated	5
people reported	5
high coverage	5
predicting depression	5
enforcement agencies	5
media metrics	5
three stimuli	5
using topic	5
coronavirus pandemic	5
strong correlation	5
sharing false	5
regression methods	5
noisy data	5
two ways	5
reputable news	5
behavioral analysis	5
age group	5
psychologically meaningful	5
research trends	5
dark red	5
specific disease	5
apache hadoop	5
black lives	5
web service	5
working hard	5
establish whether	5
nhkworld news	5
wide number	5
information extracted	5
fraudulent accounts	5
data twitter	5
also collected	5
model performance	5
i used	5
higher neuroticism	5
much higher	5
million unique	5
divert attention	5
temporal distribution	5
assess disease	5
clinical science	5
study aims	5
different hashtag	5
linguistic score	5
responsible agents	5
final dataset	5
sentiment trends	5
data streams	5
new cases	5
covid pandemic	5
information conflict	5
bayes classifier	5
lives matter	5
foreign policy	5
find words	5
final phq	5
collecting data	5
community groups	5
bipartite networks	5
two major	5
territorial police	5
twitter handles	5
societal impact	5
new research	5
based applications	5
endoscopic procedures	5
statistical significance	5
surveillance using	5
negative sentiments	5
different contexts	5
regular expressions	5
tweet id	5
violet red	5
lock downs	5
resulting network	5
first stage	5
training datasets	5
word list	5
healthmap project	5
special characters	5
jacklopresti status	5
model outperforms	5
qualitative sample	5
dbpedia spotlight	5
blue line	5
information spreading	5
agreeable people	5
will allow	5
verified accounts	5
upper endoscopy	5
home save	5
stop word	5
phq stress	5
topics discussed	5
detect depression	5
recent tweets	5
response rate	5
body ache	5
two independent	5
several tweets	5
section presents	5
strong floor	5
george floyd	5
covid twitter	5
new data	5
notation description	5
calculated using	5
forward layer	5
generally speaking	5
tweets originating	5
make use	5
computational propaganda	5
twitter platform	5
word pool	5
higher abuse	5
computer science	5
extract latent	4
uniform manifold	4
centrality measures	4
spanish government	4
topic groups	4
death toll	4
st century	4
sentiment quantification	4
liminal hotspot	4
clean data	4
take place	4
relevant knowledge	4
manually annotated	4
relevant features	4
different parties	4
collect tweets	4
second stage	4
people may	4
twitter will	4
weighted average	4
topics using	4
planned analysis	4
classification algorithm	4
first one	4
richard burgon	4
significant variation	4
high phq	4
panic production	4
data fusion	4
activity regarding	4
mental stress	4
dom cummings	4
word representations	4
media usage	4
th june	4
utc format	4
informal outbreak	4
general location	4
social sensor	4
seasonal offensive	4
disease monitoring	4
younger people	4
epidemic model	4
twitter abuse	4
subjective words	4
large numbers	4
using weakly	4
tweet body	4
cyberbullying detection	4
times larger	4
twitter search	4
random sample	4
time difference	4
emergency response	4
will continue	4
infection rate	4
coherence scores	4
twitter website	4
difficult task	4
police officers	4
essential goods	4
visual features	4
three time	4
intelligence network	4
news updates	4
false positive	4
timeline chart	4
spearman correlation	4
meeting attendance	4
muslim community	4
tweet phq	4
works well	4
research evaluation	4
global disease	4
comparative study	4
analysis shows	4
disease prevention	4
evaluation metrics	4
users worldwide	4
similar results	4
engagement scores	4
factual content	4
detect depressed	4
public life	4
group polarization	4
least three	4
wing oriented	4
information cascade	4
useful knowledge	4
critical care	4
point detection	4
take advantage	4
prediction model	4
keir starmer	4
material known	4
one example	4
hemodynamic instability	4
twitter application	4
retweeting activity	4
various altmetric	4
studies conducted	4
users expressed	4
information exchange	4
target sample	4
epidemic analysis	4
action word	4
topic distribution	4
clinical characteristics	4
blog posts	4
chinese information	4
distributed systems	4
data frameworks	4
local communities	4
susceptible nodes	4
political debate	4
many news	4
specific social	4
statistical techniques	4
toward british	4
national covid	4
shared one	4
last column	4
presidential elections	4
account creation	4
engagement score	4
main dependent	4
lagged variables	4
consensus information	4
tweets shared	4
batch size	4
top influencers	4
conversation around	4
current situation	4
tracking social	4
collecting twitter	4
haiti cholera	4
significant negative	4
marketing campaigns	4
disease alert	4
coverage across	4
one order	4
aast meeting	4
twitter developers	4
language understanding	4
different areas	4
right time	4
political communities	4
lombardy region	4
less abusive	4
test hypotheses	4
authors propose	4
biocaster ontology	4
world leaders	4
eating disorder	4
max function	4
following research	4
prime recommendations	4
whether messages	4
two pairs	4
data repository	4
news stop	4
public may	4
positive sentiment	4
three events	4
social computing	4
twitter discussions	4
social terms	4
binomial distribution	4
symptoms reported	4
related replies	4
low recall	4
mentioned previously	4
wos papers	4
review board	4
traditional media	4
word representation	4
social attention	4
moc peaks	4
strongest predictor	4
three general	4
extract features	4
economic modeling	4
also shows	4
russian information	4
web pages	4
western europe	4
selected publications	4
allowing users	4
many states	4
academic promotion	4
root tweet	4
media may	4
hundred thousand	4
individual differences	4
analysis system	4
conspiracy narrative	4
cognitive ability	4
every minute	4
recurrent unit	4
certain level	4
epidemic crisis	4
social interaction	4
research response	4
patients presenting	4
different aspects	4
comparative assessment	4
related phrases	4
related conspiracy	4
research purposes	4
two distinct	4
uniquely identifying	4
incubation period	4
large data	4
american association	4
observation period	4
change point	4
graph shows	4
directed graph	4
cascade prediction	4
followers reflect	4
science publications	4
streaming data	4
us sample	4
health concern	4
moc timeline	4
provides two	4
assessment score	4
arg pr	4
stage least	4
completed online	4
spatiotemporal patterns	4
tweet sentiment	4
identify relevant	4
economic behavior	4
death statistics	4
three columns	4
assessment method	4
mps via	4
users also	4
conceptual framework	4
geographic distribution	4
relevant keywords	4
document clustering	4
alert map	4
social graph	4
health services	4
political information	4
country i	4
feature set	4
potentially threatening	4
cybernetic feedback	4
country basis	4
various features	4
boosting immunity	4
using multi	4
runs per	4
model explained	4
flow betweenness	4
data generated	4
los angeles	4
future disaster	4
twitter allows	4
group users	4
tweet pharse	4
matthancock status	4
web search	4
strongly correlated	4
politics frame	4
general diseases	4
word analysis	4
current tweet	4
substantive sections	4
using logistic	4
information within	4
infection tweets	4
highest correlation	4
quarantine zone	4
simultaneously test	4
per game	4
naturalistic setting	4
possible future	4
done using	4
kivu ebola	4
next stage	4
tweets corresponding	4
twitter predicting	4
upper gastrointestinal	4
tell us	4
network modeling	4
digital technologies	4
json objects	4
information systems	4
extracted features	4
generated data	4
death rates	4
data preprocessing	4
chronological order	4
participant attitudes	4
sharing behaviour	4
altmetric indicators	4
first years	4
hashtag cloud	4
hierarchical clustering	4
based topic	4
supervised topic	4
learning interface	4
automated text	4
covid monitor	4
positive correlation	4
examined using	4
exogenous signal	4
national government	4
news online	4
processing algorithms	4
decreased significantly	4
covid retweeting	4
present results	4
chatter dataset	4
acceptable reliability	4
collective emotions	4
media companies	4
publication years	4
linguistic variables	4
partially mediates	4
cultural studies	4
uk law	4
family members	4
suspicious posts	4
emerging diseases	4
rapid spread	4
three dimensions	4
online communities	4
hashtags used	4
unidad ciudadana	4
related topics	4
unknowingly shared	4
apache spark	4
san francisco	4
classification task	4
weighted edges	4
published online	4
northern italy	4
suspicious accounts	4
unverified accounts	4
tweets matching	4
shifting users	4
connection strength	4
timeline trends	4
election campaign	4
highly skewed	4
rsn contracts	4
generated using	4
particular place	4
manifold approximation	4
positive negative	4
arabic content	4
distributed file	4
matter protests	4
uk police	4
academic research	4
short time	4
opposite direction	4
team offense	4
analyzing social	4
second period	4
time segments	4
pandemic intensifies	4
official twitter	4
semantic information	4
response patterns	4
another example	4
health agency	4
tweet corpus	4
google flu	4
competing interests	4
real people	4
strongly negative	4
occurrence network	4
level predictors	4
top five	4
model based	4
dynamic settings	4
equation modeling	4
fake transitions	4
majority tweet	4
vector machines	4
two data	4
timeline posts	4
time interval	4
series wins	4
physical distancing	4
user account	4
south american	4
new technologies	4
abuse towards	4
will report	4
economic impact	4
emergency alert	4
overall presence	4
will focus	4
business intelligence	4
posts per	4
topic analysis	4
i will	4
influenza surveillance	4
large groups	4
exploratory data	4
last one	4
political preference	4
among gastroenterologists	4
search queries	4
directed network	4
persuasive topics	4
table iv	4
altmetric attention	4
wide range	4
popular social	4
systematic literature	4
networks using	4
significant predictors	4
police use	4
global pandemics	4
health intelligence	4
word vector	4
user profiles	4
also includes	4
tools used	4
six different	4
using word	4
database called	4
logistic regressions	4
user behavior	4
crisis response	4
modus operandi	4
daily distribution	4
main contributions	4
study focuses	4
social restrictions	4
research works	4
means clustering	4
north american	4
proposed corexq	4
rapid research	4
early stage	4
data screening	4
visual abstracts	4
hurdle model	4
semantically meaningful	4
unannotated corpus	4
personal infection	4
value creation	4
international collaboration	4
two groups	4
louvain method	4
three primary	4
million population	4
results reported	4
raw tweets	4
uk government	4
media will	4
open data	4
also find	4
much lower	4
survey data	4
political affiliation	4
com matthancock	4
shared false	4
large datasets	4
valuable data	4
subjective terms	4
different issues	4
mr johnson	4
television contracts	4
among individuals	4
first half	4
uk sample	4
current context	4
either male	4
academic medicine	4
graph machine	4
main italian	4
structural equation	4
json format	4
attendance increases	4
research papers	4
qualitatively understand	4
influential accounts	4
privacy issues	4
national lockdown	4
using facebook	4
provide insight	4
high frequency	4
ten posts	4
attend games	4
disaster preparedness	4
media analysis	4
new zealand	4
detection system	4
unique opportunity	4
ssh publications	4
reply tweet	4
social practices	4
eight news	4
preferred topics	4
events detected	4
examine diffusion	4
gastroenterologists regarding	4
following sections	4
study twitter	4
major concerns	4
twitter chatter	4
contextual information	4
learning may	4
thousand population	4
gastrointestinal bleeding	4
analyse data	4
contain abusive	4
section reviews	4
deep state	4
retweet networks	4
abusive responses	4
embedding models	4
natural hazards	4
death score	4
media information	4
hashtags mentioned	4
mining toolkit	4
brand loyalty	4
usage analysis	4
health professions	4
feature importance	4
using commonly	4
study revealed	4
absolutely right	4
batch mode	4
marketing measures	4
suspicious profiles	4
user activity	4
travel north	4
economic disruption	4
sports teams	4
received abusive	4
projection procedure	4
political figureheads	4
authors found	4
based clustering	4
twitter social	4
system continuously	4
sans frontieres	4
independent ols	4
clinical oncology	4
depressed user	4
depth analysis	4
data related	4
count data	4
healthcare providers	4
health specialists	4
public officials	4
suspended accounts	4
empirical analysis	4
phq category	4
online conversations	4
valuable insight	4
center right	4
hate tweets	4
important news	4
information associated	4
key terms	4
video comments	4
feature fusion	4
surrounding covid	4
negative emoticon	4
proposed model	4
data point	4
publications across	4
linguistic inquiry	4
general awareness	4
convolutional neural	4
floor effect	4
users express	4
across five	4
change across	4
ebola epidemic	4
higher numbers	4
pse publications	4
cascade length	4
thousand tweets	4
also identified	4
reliable data	4
recent advances	4
collection process	4
used hashtags	4
unprecedented times	4
official sources	4
tweet using	4
tweets expressing	4
time disease	4
labour party	4
spatiotemporal scale	4
node degree	4
media communications	4
systems based	4
hateful tweet	4
highly contagious	4
chat platform	4
syndromic surveillance	4
persuasive arguments	4
learning algorithm	4
television viewership	4
topic classification	4
many parents	4
information source	4
baseline methods	4
urgent scenarios	4
public risk	4
measured using	4
research findings	4
white women	4
disinformation messages	4
seven days	4
twitter json	4
buying group	4
another one	4
used data	4
time using	4
biomedical text	4
latent topics	4
useful information	4
individual tweet	4
louvain algorithm	4
health monitoring	4
first public	4
global health	4
first place	4
previous results	4
target audiences	4
institutional review	4
problems related	4
network features	4
digital data	4
epidemiological indicators	4
best performance	4
manually labeled	4
worldwide data	4
trending covid	4
english words	4
seed labels	4
see eq	4
municipal government	4
ongoing covid	4
association rules	4
table summarizes	4
available tweets	4
far greater	4
concerned covid	4
classification methods	4
prior research	4
quick search	4
mediation results	4
respondents indicated	4
media tweet	4
patient privacy	4
analytics tools	4
resource center	4
sentiment strength	4
independent cascade	4
topics identified	4
interactive web	4
linguistic word	4
search function	4
health professionals	4
tv programs	4
given inclusion	4
may use	4
respondents thought	4
brief summary	4
open scientific	4
different domains	4
cosine similarity	4
wide variety	4
daily deaths	4
similar trend	4
support system	4
specific content	4
semantic lexicons	4
important role	4
online misinformation	4
centered around	4
scalable seeded	4
events may	4
traditional data	4
human trafficking	4
generated topics	4
classification approach	4
italian prime	4
consensus among	4
leibler divergence	4
model parameters	4
shed light	4
monitor austria	4
regarding covid	4
three days	4
story source	4
three stories	4
news article	4
computational complexity	4
migration policing	4
manual coding	4
study news	4
twitter hashtags	4
threatening media	4
crime specific	4
moderately correlated	4
five main	4
instructional information	4
retweeting time	4
first author	4
slate blue	4
political orientations	4
police force	4
strongly positive	4
cnn network	4
document corpus	4
different levels	4
phq label	4
epidemic events	4
research paper	4
tweet type	4
abuse received	4
linguistic scores	4
makes sense	4
high risk	4
coronavirus conspiracy	4
allow us	4
situational information	4
critical public	4
commons attribution	4
democratic party	4
human activities	4
pos tags	4
actual case	4
warning signs	3
mlb team	3
including social	3
conference material	3
behavior via	3
frequency inverse	3
among people	3
examples include	3
analysis will	3
specific keywords	3
information system	3
internet research	3
trust authorities	3
may contain	3
various tasks	3
british political	3
trained model	3
bandwidth usage	3
crime type	3
similar trends	3
retweeting messages	3
pause censorship	3
analytics approach	3
epub ahead	3
negatively correlated	3
common covid	3
spatial context	3
network obtained	3
misinformation epidemic	3
following analysis	3
gastroenterologists agreed	3
processing methods	3
methodology exactly	3
time twitter	3
national linguistic	3
risk perceptions	3
words present	3
partido justicialista	3
certain political	3
convolution layer	3
numbing effect	3
academic institutions	3
also refer	3
relevant news	3
direct relationship	3
related knowledge	3
chicago cubs	3
various types	3
research model	3
print media	3
distancing policies	3
spread research	3
independent variables	3
vary significantly	3
web interface	3
politicians receiving	3
rich set	3
publications mentioned	3
distributed across	3
conference materials	3
among various	3
decision trees	3
large dataset	3
across publications	3
binary segmentation	3
public support	3
repeated exposure	3
van prooijen	3
although twitter	3
raw count	3
disease community	3
frame included	3
also rallied	3
regular user	3
systematic mapping	3
closures group	3
similarity measures	3
recognition software	3
endoscopic interventions	3
feature space	3
top bigrams	3
moral contests	3
received relatively	3
response time	3
score greater	3
users according	3
conventional media	3
feb th	3
across time	3
methodology used	3
network featured	3
tweet based	3
user behaviours	3
labeling approach	3
binary accuracy	3
section iii	3
two variables	3
data dictionary	3
size conferred	3
become essential	3
enabled us	3
political analysis	3
gastroenterology fellows	3
hospital beds	3
media disinformation	3
north korea	3
one side	3
learning classifier	3
introductory paragraph	3
web services	3
fraudulent financial	3
medical students	3
table illustrates	3
overall pattern	3
complementary data	3
law model	3
recent analysis	3
computational model	3
normal users	3
higher accuracy	3
spatial decision	3
social platform	3
local health	3
topics across	3
predictor set	3
hybrid approach	3
outbreak classes	3
related keywords	3
government decision	3
found containing	3
conservatism scale	3
best topic	3
identify shifting	3
right quadrant	3
usage graph	3
geographical region	3
fan motivations	3
research will	3
contour lines	3
il fatto	3
may change	3
current research	3
epidemic research	3
dataset contains	3
knowledge base	3
media analytics	3
collective alarm	3
study based	3
encountered clinical	3
structured data	3
least two	3
disk i	3
social influence	3
may require	3
maximum number	3
link increased	3
two indicators	3
available datasets	3
hadoop distributed	3
supervised learning	3
provide insights	3
positive samples	3
good balance	3
study will	3
status id	3
media provides	3
specific subject	3
static prediction	3
information operation	3
examined whether	3
unique tweets	3
varies across	3
entity instance	3
transition point	3
news data	3
location data	3
cambridge analytica	3
bot activity	3
network graph	3
noun phrases	3
prominent politicians	3
features extracted	3
emerging topics	3
like facebook	3
also received	3
either positive	3
continuing medical	3
mps using	3
main stages	3
susceptible agents	3
time intervals	3
cent army	3
containing urls	3
massive relevance	3
general population	3
billion tweets	3
time credit	3
intensively use	3
relevant words	3