1. Analyzing structure vs analyzing content Short sidestep: Agenda setting and Framing Structure Content Conc
Fundamentals of Data Science: Case “Political
Communication”
Damian Trilling
d.c.trilling@uva.nl
@damian0604
www.damiantrilling.net
Afdeling Communicatiewetenschap
Universiteit van Amsterdam
19-09-2016
Fundamentals of Data Science: Case “Political Communication” Damian Trilling
2. Analyzing structure vs analyzing content Short sidestep: Agenda setting and Framing Structure Content Conc
Last week
1 some themes in political communication research
• polarization
• fragmentation
• and the way politicans use social media
2 Twitter API, preprocessing, geodata, sentiment analysis
Fundamentals of Data Science: Case “Political Communication” Damian Trilling
3. Analyzing structure vs analyzing content Short sidestep: Agenda setting and Framing Structure Content Conc
This week
Digging deeper into the content of the tweets
Fundamentals of Data Science: Case “Political Communication” Damian Trilling
4. Analyzing structure vs analyzing content Short sidestep: Agenda setting and Framing Structure Content Conc
Today
1 Analyzing structure vs analyzing content
2 Short sidestep: Agenda setting and Framing
3 Studies that analyze structure of the Twittersphere
4 Studies that analyze content of tweets
Issues
Responses to TV debates
Incivility
5 Conclusion
Fundamentals of Data Science: Case “Political Communication” Damian Trilling
5. Analyzing structure vs analyzing content Short sidestep: Agenda setting and Framing Structure Content Conc
Structure
Analyzing structure vs analyzing content
Fundamentals of Data Science: Case “Political Communication” Damian Trilling
6. Analyzing structure vs analyzing content Short sidestep: Agenda setting and Framing Structure Content Conc
Structure
Analyzing Twitter data
Analyzing the structure
• Number of Tweets over time
• singleton/retweet ratio
• Distribution of number of Tweets per user
• Interaction networks
Bruns, A., & Stieglitz, S. (2013). Toward more systematic Twitter analysis: Metrics for tweeting activities.
International Journal of Social Research Methodology. doi:10.1080/13645579.2012.756095
Fundamentals of Data Science: Case “Political Communication” Damian Trilling
7. Analyzing structure vs analyzing content Short sidestep: Agenda setting and Framing Structure Content Conc
Structure
Analyzing Twitter data
Analyzing the structure
• Number of Tweets over time
• singleton/retweet ratio
• Distribution of number of Tweets per user
• Interaction networks
⇒ Focus on the amount of content and on the question who
interacts with whom, not on what is said
Bruns, A., & Stieglitz, S. (2013). Toward more systematic Twitter analysis: Metrics for tweeting activities.
International Journal of Social Research Methodology. doi:10.1080/13645579.2012.756095
Fundamentals of Data Science: Case “Political Communication” Damian Trilling
8. Analyzing structure vs analyzing content Short sidestep: Agenda setting and Framing Structure Content Conc
Content
Analyzing Twitter data
Analyzing the content
• Sentiment analysis
• Word frequencies
• regexp searches
• Word cooccurrences (⇒topics, frames)
• co-occurrence networks
• PCA
• LDA
• . . .
Fundamentals of Data Science: Case “Political Communication” Damian Trilling
9. Analyzing structure vs analyzing content Short sidestep: Agenda setting and Framing Structure Content Conc
Content
Analyzing Twitter data
Analyzing the content
• Sentiment analysis
• Word frequencies
• regexp searches
• Word cooccurrences (⇒topics, frames)
• co-occurrence networks
• PCA
• LDA
• . . .
⇒ Focus on what is said
Fundamentals of Data Science: Case “Political Communication” Damian Trilling
10. Analyzing structure vs analyzing content Short sidestep: Agenda setting and Framing Structure Content Conc
Content
Systematizing analytical approaches
⇒ It depends on your reserach question which approach is
more interesting!
Fundamentals of Data Science: Case “Political Communication” Damian Trilling
11. Analyzing structure vs analyzing content Short sidestep: Agenda setting and Framing Structure Content Conc
Content
Systematizing analytical approaches
⇒ It depends on your reserach question which approach is
more interesting!
But probably the most interesting thing is to combine them
both
Fundamentals of Data Science: Case “Political Communication” Damian Trilling
13. Analyzing structure vs analyzing content Short sidestep: Agenda setting and Framing Structure Content Conc
Agenda setting
Beyond simplistic stimulus-response models of media effects:
Media effects are not so much about
how we think, but what we think
aboutMcCombs, M, & Shaw, D (1972). The agenda-setting function of mass media. Public Opinion Quarterly, 36: 176.
doi:10.1086/267990
Fundamentals of Data Science: Case “Political Communication” Damian Trilling
14. Analyzing structure vs analyzing content Short sidestep: Agenda setting and Framing Structure Content Conc
Framing
“To frame is to select some aspects of a perceived reality and
make them more salient in a communicating text, in such a way as
to promote a particular problem definition, causal interpretation,
moral evaluation, and/or treatment recommendation for the item
described”
Entman, R. M. (1993). Framing: Toward clarification of a fractured paradigm. Journal of Communication, 43,
51–58. doi:10.1111/j.1460-2466.1993.tb01304.x
Fundamentals of Data Science: Case “Political Communication” Damian Trilling
16. Analyzing structure vs analyzing content Short sidestep: Agenda setting and Framing Structure Content Conc
The Twittersphere
Mapping the Austrian Twittersphere
• How do politicians, journalists, and citizens interact?
• How do topics between news coverage and tweets overlap?
(already content)
Ausserhofer, J., & Maireder, A. (2013). National Politics on Twitter. Information, Communication & Society,
16(3), 291—314. doi:10.1080/1369118X
Fundamentals of Data Science: Case “Political Communication” Damian Trilling
17.
18.
19. “In general, famous journalists, experts and politicians are central
actors within the Austrian political Twittersphere and form their
own, dense and influential subnetwork within the broader sphere.
Non-professionals may participate in this network, provided that
they engage receptive members of the elite who act as ‘bridges’
between subnetworks. However, when the discussion involves
certain topics, niche authorities emerge, and these authorities –
including a few left-wing activists and bloggers – join other
political professionals as central information hubs.”
Ausserhofer & Maireder 2013, p. 19
20.
21. “While topics such as the financial crisis were massively
represented in the newspapers and on TV, hardly anyone tweeted
about such topics on Twitter. A similar phenomenon could be
observed with the ongoing coverage of corruption-related
investigations, about which only a few users bothered to tweet.
Short-living topics such as the aforementioned ball of the
right-wing fraternities and the squatting of an abandoned house
and the forced eviction of its ‘residents’ were popular topics on
Twitter. A further explanation of why these topics are more
popular on Twitter than in mass media is that activists use the
service not only to discuss but also to facilitate their activities.”
Ausserhofer & Maireder 2013, p. 19
23. Analyzing structure vs analyzing content Short sidestep: Agenda setting and Framing Structure Content Conc
Issues
Networks of issues
• Which topics are co-mentioned by the same users?
• Which topics are co-mentioned by different types of accounts?
Vargo, C. J., Guo, L., McCombs, M., & Shaw, D. L. (2014). Network Issue Agendas on Twitter During the 2012
U.S. Presidential Election. Journal of Communication, 64, 296–316. doi:10.1111/jcom.12089
Fundamentals of Data Science: Case “Political Communication” Damian Trilling
24. Analyzing structure vs analyzing content Short sidestep: Agenda setting and Framing Structure Content Conc
Issues
Networks of issues
• Which topics are co-mentioned by the same users?
• Which topics are co-mentioned by different types of accounts?
• Sentistrength + in combination with Obama/Romney to
determine who supports whom
• simple keyword searches (dictionary-approach) for topic
classification
• network analysis
Vargo, C. J., Guo, L., McCombs, M., & Shaw, D. L. (2014). Network Issue Agendas on Twitter During the 2012
U.S. Presidential Election. Journal of Communication, 64, 296–316. doi:10.1111/jcom.12089
Fundamentals of Data Science: Case “Political Communication” Damian Trilling
25. Analyzing structure vs analyzing content Short sidestep: Agenda setting and Framing Structure Content Conc
Issues
Networks of issues
• General-interest media issue network predicts issue network of
Obama supporters
• Partisan media issue network predicts issue network of
Romney supporters
Vargo, C. J., Guo, L., McCombs, M., & Shaw, D. L. (2014). Network Issue Agendas on Twitter During the 2012
U.S. Presidential Election. Journal of Communication, 64, 296–316. doi:10.1111/jcom.12089
Fundamentals of Data Science: Case “Political Communication” Damian Trilling
26. Analyzing structure vs analyzing content Short sidestep: Agenda setting and Framing Structure Content Conc
Responses to TV debates
Second Screen
• Linking events to Twitter reactions
• Linking candidate behavior to Twitter reactions
Vergeer, M., & Franses, P. H. (2015). Live audience responses to live televised election debates: Time series
analysis of issue salience and party salience on audience behavior. Information, Communication & Society
doi:10.1080/1369118X.2015.1093526
Trilling, D. (2015). Two different debates? Investigating the relationship between a political debate on TV and
simultaneous comments on Twitter. Social Science Computer Review, 33(3), 259–276.
doi:10.1177/0894439314537886
Yıldırım, A., Üsküdarlı, S., & Özgür, A. (2016). Identifying Topics in Microblogs Using Wikipedia. Plos One,
11(3), e0151885. doi:10.1371/journal.pone.0151885
Fundamentals of Data Science: Case “Political Communication” Damian Trilling
27. Analyzing structure vs analyzing content Short sidestep: Agenda setting and Framing Structure Content Conc
Responses to TV debates
Second Screen
• Linking events to Twitter reactions
• Linking candidate behavior to Twitter reactions
Central question
How do people react to TV debates?
Vergeer, M., & Franses, P. H. (2015). Live audience responses to live televised election debates: Time series
analysis of issue salience and party salience on audience behavior. Information, Communication & Society
doi:10.1080/1369118X.2015.1093526
Trilling, D. (2015). Two different debates? Investigating the relationship between a political debate on TV and
simultaneous comments on Twitter. Social Science Computer Review, 33(3), 259–276.
doi:10.1177/0894439314537886
Yıldırım, A., Üsküdarlı, S., & Özgür, A. (2016). Identifying Topics in Microblogs Using Wikipedia. Plos One, 11(3),
e0151885. doi:10.1371/journal.pone.0151885
Fundamentals of Data Science: Case “Political Communication” Damian Trilling
28. Example 1:
relating word frequencies to each other
Trilling, D. (2015). Two different debates? Investigating the relationship between a political debate on TV and
simultaneous comments on Twitter. Social Science Computer Review, 33(3), 259–276.
doi:10.1177/0894439314537886
29.
30.
31.
32. Analyzing structure vs analyzing content Short sidestep: Agenda setting and Framing Structure Content Conc
Responses to TV debates
A way of visualizing this
font size ∼ relative frequency within copus
distance to y-axis ∼ log-likelikelihood (= difference between corpora)
Fundamentals of Data Science: Case “Political Communication” Damian Trilling
33. Example 2:
manually classify most frequent terms into topics,
subsequent time series analysis
Vergeer, M., & Franses, P. H. (2015). Live audience responses to live televised election debates: Time series
analysis of issue salience and party salience on audience behavior. Information, Communication & Society.
doi:10.1080/1369118X.2015.1093526
34.
35.
36. Example 3:
Using external datasource (wikipedia) for topic classification
Yıldırım, A., Üsküdarlı, S., & Özgür, A. (2016). Identifying Topics in Microblogs Using Wikipedia. Plos One, 11(3),
e0151885. doi:10.1371/journal.pone.0151885
37.
38.
39. Analyzing structure vs analyzing content Short sidestep: Agenda setting and Framing Structure Content Conc
Incivility
Incivility
Who uses incivil language on Twitter?
Vargo, C. J., & Hopp, T. (2015). Socioeconomic status, social capital, and partisan polarity as predictors of
political incivility on Twitter: A congressional district-level analysis. Social Science Computer Review
doi:10.1177/0894439315602858
Fundamentals of Data Science: Case “Political Communication” Damian Trilling
40. Analyzing structure vs analyzing content Short sidestep: Agenda setting and Framing Structure Content Conc
Incivility
Incivility
Who uses incivil language on Twitter?
incivility
(1) name-calling; (2) threats; (3) vulgarities; (4) abusive or foul
language; (5) xenophobia; (6) hateful language, epithets, or slurs;
(7) racist or bigoted sentiments; (8) disparaging comments on the
basis of race/ethnicity; and (9) use of stereotypes
Vargo, C. J., & Hopp, T. (2015). Socioeconomic status, social capital, and partisan polarity as predictors of
political incivility on Twitter: A congressional district-level analysis. Social Science Computer Review
doi:10.1177/0894439315602858
Fundamentals of Data Science: Case “Political Communication” Damian Trilling
41. Analyzing structure vs analyzing content Short sidestep: Agenda setting and Framing Structure Content Conc
Incivility
Incivility
Who uses incivil language on Twitter?
incivility
(1) name-calling; (2) threats; (3) vulgarities; (4) abusive or foul
language; (5) xenophobia; (6) hateful language, epithets, or slurs;
(7) racist or bigoted sentiments; (8) disparaging comments on the
basis of race/ethnicity; and (9) use of stereotypes
dictionary approach, based on existing word lists
Vargo, C. J., & Hopp, T. (2015). Socioeconomic status, social capital, and partisan polarity as predictors of
political incivility on Twitter: A congressional district-level analysis. Social Science Computer Review
doi:10.1177/0894439315602858
Fundamentals of Data Science: Case “Political Communication” Damian Trilling
42. Analyzing structure vs analyzing content Short sidestep: Agenda setting and Framing Structure Content Conc
Incivility
Incivility
The central question
Do factors that are thought to be indicators of a functioning
democratic discourse (like low polarization) translate to a civil
discourse on social media?
Vargo, C. J., & Hopp, T. (2015). Socioeconomic status, social capital, and partisan polarity as predictors of
political incivility on Twitter: A congressional district-level analysis. Social Science Computer Review
doi:10.1177/0894439315602858
Fundamentals of Data Science: Case “Political Communication” Damian Trilling
43.
44. “Our results suggested that uncivil discourse was highest in
districts that were characterized, in part, by factors traditionally
thought to be indicative of a healthy and diverse democracy (i.e.,
low levels of partisan polarity and high levels of racial diversity).”
“Notably, we failed to either fully or partially support a number of
our hypotheses.”
Vargo & Hopp, 2015, p. 17
45. “A number of limitations temper the present findings. First, the
nature of the data severely limits the generalizability of our
findings. The source of data here, Twitter, is, at best, an
instantaneous measure of behavior, not a durable measure of
emotion or feelings (Vieweg, 2010). Moreover, Twitter cannot be
reasonably understood to be a directly reliable proxy for public
opinion in general. Also, the corpus here was limited to a specific
event, the 2012 general election. The messages gathered in this
analysis were also directed at a specific political candidate (e.g.,
Obama and/or Romney). While the findings still yield important
conclusions toward discourse, democracy, and general elections, we
cannot use the current results to make generalizations about the
state of political discussion as a whole (either on or off of Twitter).”
Vargo & Hopp, 2015, p. 17
46. Remember:
This was just a tiny selection to give you some inspiration
about what one can research.
There are a bunch of other interesting studies and approaches.
47. Analyzing structure vs analyzing content Short sidestep: Agenda setting and Framing Structure Content Conc
Further reading
Jungherr, A. (2016). Twitter use in election campaigns: A
systematic literature review. Journal of Information Technology &
Politics, 13(1), 72–91. doi:10.1080/19331681.2015.1132401
Fundamentals of Data Science: Case “Political Communication” Damian Trilling
48. Analyzing structure vs analyzing content Short sidestep: Agenda setting and Framing Structure Content Conc
Questions?
d.c.trilling@uva.nl
@damian0604
www.damiantrilling.net
Fundamentals of Data Science: Case “Political Communication” Damian Trilling