As more and more people use social media to communicate their view and perception of elections, researchers have increasingly been collecting and analyzing data from social media platforms. Our research focuses on social media communication related to the 2013 election of the German parlia-ment [translation: Bundestagswahl 2013]. We constructed several social media datasets using data from Facebook and Twitter. First, we identified the most relevant candidates (n=2,346) and checked whether they maintained social media accounts. The Facebook data was collected in November 2013 for the period of January 2009 to October 2013. On Facebook we identified 1,408 Facebook walls containing approximately 469,000 posts. Twitter data was collected between June and December 2013 finishing with the constitution of the government. On Twitter we identified 1,009 candidates and 76 other agents, for example, journalists. We estimated the number of relevant tweets to exceed eight million for the period from July 27 to September 27 alone. In this document we summarize past research in the literature, discuss possibilities for research with our data set, explain the data collection procedures, and provide a description of the data and a discussion of issues for archiving and dissemination of social media data.
Unlock Your Social Media Potential with IndianLikes - IndianLikes.com
PEP-TF: Social Media Monitoring of the Campaigns for the 2013 German Bundestag Elections on Facebook and Twitter
1. PEP-TF
Social Media Monitoring of the Campaigns for the 2013
German Bundestag Elections on Facebook and Twitter
Lead: Lars Kaczmirek1, Philipp Mayr1, Ravi Vatrapu*
Members: Arnim Bleier1, Manuela Blumenberg1, Tobias Gummer1, Kjeld
Hansen*, Abid Hussein*, Zeshan Ali Jaffari*, Katharina Kinder-Kurlanda1, Haiko
Lietz, Markus Strohmaier1, Claudia Wagner1, Katrin Weller1, Christof
Wolf1, Chris James Zimmerman*
1 GESIS – Leibniz Institute for the Social Sciences
* Department of IT Management, Copenhagen Business School
Please cite as:
Lars Kaczmirek, Philipp Mayr, Ravi Vatrapu, Arnim Bleier, Manuela Blumenberg, Tobias Gummer, Kjeld
Hansen, Abid Hussein, Zeshan Ali Jaffari, Katharina Kinder-Kurlanda, Haiko Lietz, Markus
Strohmaier, Claudia Wagner, Katrin Weller, Christof Wolf, Chris James Zimmerman. (2014). PEP-TF: Social
Media Monitoring of the Campaigns for the 2013 German Bundestag Elections on Facebook and Twitter.
Presentation at the Social Media Communication Workshop, May 28-29, 2014, MZES, Germany.
2. Research domain
• To examine communication structures in online media and how such
data can add new insights in comparison to existing data from surveys,
print media analyses, and the TV candidates’ debate.
• Describe and visualize dynamics in political debates in socia media: e.g.
pure social media phenomena like hypes, memes and hoaxes
• Analyze the influence of classical media on social media (e.g. second
screen analyses)
2
3. Social Media Data
political communication with respect to the German Bundestag election on
22 Sep 2013
CBS
Facebook
WTS
Twitter
DBG
Sample
3
5. Working Paper:
Social Media Monitoring of the Campaigns for the 2013
German Bundestag Elections on Facebook and Twitter
Lars Kaczmirek1, Philipp Mayr1, Ravi Vatrapu*, Arnim
Bleier1, Manuela Blumenberg1, Tobias Gummer1, Abid
Hussein*, Katharina Kinder-Kurlanda1, Kaveh Manshaei1, Mark
Thamm1, Katrin Weller1, Alexander Wenz1, Christof Wolf1
1 GESIS – Leibniz Institute for the Social Sciences
* Department of IT Management, Copenhagen Business School
6. PEP-TF Fact Sheet
• 2346 candidates
• 1408 Facebook walls with 469 000 posts
• 1009 Twitter accounts with >8 million tweets
– +76 gate keepers, such as journalists
– +100 authorities
• 175 topics and campaign hashtags, #btw13 #partei #hartz4 #maut …
• 20 hashtags about #nsa #snowden #yeswescan …
• 67% of the candidates have a wall or page,
• 43% have a Twitter account
6
7. Corpora
1. Facebook corpus of candidates
2. Twitter corpus of candidates
3. Twitter corpus of media agents
4. Twitter hashtag corpus of basic political topics
5. Twitter hashtag corpus of media topics
6. Twitter hashtag corpus about NSA / Snowden
Data collection periods:
– Twitter: 20 June 2013 - 17 December 2013
– Facebook: in November 2013 for the period of
1 Jan 2009 – 31 Oct 2013
7
@
@
#
#
#
C
8. C1 + C2: retrieving the Bundestag candidates
• Reasonable likelihood of becoming a member of the Bundestag
• Stage 1
– Search listings of candidatures
– Web, email and telephone call
– Continuous update and extension necessary
• Stage 2
– Identify and verify Facebook and Twitter accounts
– Verification
1. Is a reference to the party, for example a party logo visible? Are Facebook
friends and Twitter followers members of this party?
2. Do the candidate’s personal or party website link to the profile?
3. Can the candidate be recognized via image or constituency (for direct
candidates)?
4. Accounts used only for private purposes were not included.
8
@
9. Data types
9
Attribute Description
guid Unique ID to identify an object, such as the account name
Date and time stamps Identify the date and time of an action, such as a like
Comment Information about a comment on the wall
Likes Information about a like on the wall
Post texts Information about the posted text
Actor attributes Information about the person who made an action, such as a like
10. Data types
10
Attribute Description
_id tweet ID
userid numeric user ID
screenName alpha numeric user ID
name username
replyid ID of the original tweet, if this tweet is a reply; otherwise -1
replyuserid user ID of original tweet
createdAt date of tweet
tweettext text of this tweet
11. Number of tweets per day
Note. Green peaks coincide with candidates TV debate and election day.
11
12. Corpus 2: Frequency of Tweets over time
Absolute
frequency of
tweets is
correlated with
major political
events.
Intensity was
highest in the
mid & end of
campaigning.
Gummer et al. (2014)
13. Corpus 2: Frequency of Tweets over time by party
PIRATEN do
account for
most of the
Tweets.
Bias in
aggregated
distribution
when not
controlling
for parties.
Gummer et al. (2014)
14. When Politicians Talk
Assessing Online Conversational Practices of Political Parties on Twitter
following
14
Lietz et al. (2014)
mentioningretweeting
15. Further research agenda
• Identification of topics in short texts (140 characters) via unsupervised
topic modeling (large PEP-TF corpora)
• Identification and visualization of topic drifts over time (small PEP-TF
corpora)
• Social media data for search support in portals like sowiport
– Recommendation services for „hot“ topics, debates and persons
16. Thank you
• Gummer, T., Roßmann, J., & Wolf, C. (2014). Candidates’ Twitter Use in the
German Election 2013. Presentation at the General Online Research 2014,
Cologne, Germany.
• Lietz, H., Wagner, C., Bleier, A., & Strohmaier, M. (2014). When politicians talk:
Assessing online conversational practices of political parties on twitter. In
International AAAI Conference on Weblogs and Social Media (ICWSM2014),
Ann Arbor, MI, USA, June 2-4, 2014.
• Thamm, M., & Bleier, A. (2013). When Politicians Tweet: A Study on the
Members of the German Federal Diet. In ACM Web Science 2013. Paris,
France. Retrieved from http://www.websci13.org/
• Weller, K., Bruns, A., Burgess, J., Mahrt, M. & Puschmann, C. (Eds.). (2014).
Twitter and Society. New York: Peter Lang.
16