Other Topics You May Also Agree or Disagree: Modeling Inter-Topic Preferences using Tweets and Matrix Factorization

Proceedings of the 55th Annual Meeting of the Association for Computational Linguistics, pages 398–408 Vancouver, Canada, July 30 - August 4, 2017. c⃝2017 Association for Computational Linguistics https://doi.org/10.18653/v1/P17-1037 Proceedings of the 55th Annual Meeting of the Association for Computational Linguistics, pages 398–408 Vancouver, Canada, July 30 - August 4, 2017. c⃝2017 Association for Computational Linguistics https://doi.org/10.18653/v1/P17-1037 Other Topics You May Also Agree or Disagree: Modeling Inter-Topic Preferences using Tweets and Matrix Factorization Akira Sasaki, Kazuaki Hanawa, Naoaki Okazaki, and Kentaro Inui Graduate School of Information Sciences Tohoku University {aki-s, hanawa, okazaki, inui}@ecei.tohoku.ac.jp Abstract We present in this paper our approach for modeling inter-topic preferences of Twit- ter users: for example, those who agree with the Trans-Pacific Partnership (TPP) also agree with free trade. This kind of knowledge is useful not only for stance detection across multiple topics but also for various real-world applications includ- ing public opinion surveys, electoral pre- dictions, electoral campaigns, and online debates. In order to extract users’ prefer- ences on Twitter, we design linguistic pat- terns in which people agree and disagree about specific topics (e.g., “A is com- pletely wrong”). By applying these lin- guistic patterns to a collection of tweets, we extract statements agreeing and dis- agreeing with various topics. Inspired by previous work on item recommendation, we formalize the task of modeling inter- topic preferences as matrix factorization: representing users’ preferences as a user- topic matrix and mapping both users and topics onto a latent feature space that ab- stracts the preferences. Our experimental results demonstrate both that our proposed approach is useful in predicting missing preferences of users and that the latent vector representations of topics success- fully encode inter-topic preferences. 1 Introduction Social media have changed the way people shape public opinion. The latest survey by the Pew Research Center reported that a majority of US adults (62%) obtain news via social media, and of those, 18% do so often (Gottfried and Shearer, 2016). Given that news and opinions are shared and amplified by friend networks of individu- als (Jamieson and Cappella, 2008), individuals are thereby isolated from information that does not fit well with their opinions (Pariser, 2011). Ironi- cally, cutting-edge social media technologies pro- mote ideological groups even with its potential to deliver diverse information. A large number of studies already analyzed discussions, interactions, influences, and commu- nities on social media along the political spec- trum from liberal to conservative (Adamic and Glance, 2005; Zhou et al., 2011; Cohen and Ruths, 2013; Bakshy et al., 2015; Wong et al., 2016). Even though these studies provide intuitive vi- sualizations and interpretations along the liberal- conservative axis, political analysts argue that the axis is flawed and insufficient for representing public opinion and ideologies (Kerlinger, 1984; Maddox and Lilie, 1984). A potential solution for analyzing multiple axes of the political spectrum on social media is stance detection (Thomas et al., 2006; Somasundaran and Wiebe, 2009; Murakami and Raymond, 2010; Anand et al., 2011; Walker et al., 2012; Moham- mad et al., 2016; …

Content truncated to 500 words