Understanding Discourse Acts: Political Campaign Messages Classification on Facebook and Twitter

  • Feifei Zhang
  • Jennifer Stromer-GalleyEmail author
  • Sikana Tanupabrungsun
  • Yatish Hegde
  • Nancy McCracken
  • Jeff Hemsley
Conference paper
Part of the Lecture Notes in Computer Science book series (LNCS, volume 10354)


To understand political campaign messages in depth, we developed automated classification models for classifying categories of political campaign Twitter and Facebook messages, such as calls-to-action and persuasive messages. We used 2014 U.S. governor’s campaign social media messages to develop models, then tested these models on a randomly selected 2016 U.S. presidential campaign social media dataset. Our classifiers reach .75 micro-averaged F value on training sets and .76 micro-averaged F value on test sets, suggesting that the models can be applied to classify English-language political campaign social media messages. Our study also suggests that features afforded by social media help improve classification performance in social media documents.


Automated classification Political campaign Social media Supervised learning Text mining 



We thank Dr. Bei Yu’s helpful feedback on this paper. The project was supported by the Tow Center for Digital Journalism at Columbia University and the Center for Computational and Data Sciences at the School of Information Studies at Syracuse University.


  1. 1.
    Conover, M.D., Gonçalves, B., Ratkiewicz, J., Flammini, A., Menczer, F.: Predicting the political alignment of twitter users. In: Privacy, Security, Risk and Trust and 2011 IEEE Third International Conference on Social Computing, pp. 192–199. IEEE (2011)Google Scholar
  2. 2.
    Gimpel, K., Schneider, N., O’Connor, B., Das, D., Mills, D., Eisenstein, J., Heilman, M., Yogatama, D., Flanigan, J., Smith, N.A.: Part-of-speech tagging for twitter: annotation, features, and experiments. In: Proceedings of the 49th Annual Meeting of the Association for Computational Linguistics: Human Language Technologies: short papers, vol. 2, pp. 42–47. Association for Computational Linguistics (2011)Google Scholar
  3. 3.
    Hegde, Y.: fb-page-scraper: version 1.33 (2016).
  4. 4.
    Hemsley, J., Ceskavich, B., Tanupabrungsun, S.: Syracuse Social Media Collection Toolkit (2014).
  5. 5.
    Jamieson, K.H., Waldman, P., Sherr, S.: Eliminate the negative? Categories of analysis for political advertisements. In: Crowded Airwaves: Campaign Advertising in Elections, pp. 44–64 (2000)Google Scholar
  6. 6.
    Lombard, M., Snyder-Duch, J., Bracken, C.C.: Content analysis in mass communication: assessment and reporting of intercoder reliability. Hum. Commun. Res. 28(4), 587–604 (2002)CrossRefGoogle Scholar
  7. 7.
    Pedregosa, F., Varoquaux, G., Gramfort, A., Michel, V., Thirion, B., Grisel, O., Blondel, M., Prettenhofer, P., Weiss, R., Dubourg, V., Vanderplas, J.: Scikit-learn: machine learning in python. J. Mach. Learn. Res. 12, 2825–2830 (2011)MathSciNetzbMATHGoogle Scholar
  8. 8.
    Sriram, B., Fuhry, D., Demir, E., Ferhatosmanoglu, H., Demirbas, M.: Short text classification in twitter to improve information filtering. In: Proceedings of the 33rd International ACM SIGIR Conference on Research and Development in Information Retrieval, pp. 841–842. ACM (2010)Google Scholar
  9. 9.
    Van Asch, V.: Macro-and micro-averaged evaluation measures [basic draft] (2013)Google Scholar

Copyright information

© Springer International Publishing AG 2017

Authors and Affiliations

  • Feifei Zhang
    • 1
  • Jennifer Stromer-Galley
    • 1
    Email author
  • Sikana Tanupabrungsun
    • 1
  • Yatish Hegde
    • 1
  • Nancy McCracken
    • 1
  • Jeff Hemsley
    • 1
  1. 1.School of Information StudiesSyracuse UniversitySyracuseUSA

Personalised recommendations