Open Extraction of Fine-Grained Political Statements

Proceedings of the 2015 Conference on Empirical Methods in Natural Language Processing, pages 76–85, Lisbon, Portugal, 17-21 September 2015. c⃝2015 Association for Computational Linguistics. Open Extraction of Fine-Grained Political Statements David Bamman School of Information University of California, Berkeley Berkeley, CA 94720, USA dbamman@berkeley.edu Noah A. Smith Computer Science & Engineering University of Washington Seattle, WA 98195, USA nasmith@cs.washington.edu Abstract Text data has recently been used as evi- dence in estimating the political ideologies of individuals, including political elites and social media users. While inferences about people are often the intrinsic quan- tity of interest, we draw inspiration from open information extraction to identify a new task: inferring the political import of propositions like OBAMA IS A SOCIAL- IST. We present several models that ex- ploit the structure that exists between peo- ple and the assertions they make to learn latent positions of people and propositions at the same time, and we evaluate them on a novel dataset of propositions judged on a political spectrum. 1 Introduction Over the past few years, much work has fo- cussed on inferring political preferences of peo- ple from their behavior, both in unsupervised and supervised settings. Classical ideal point models (Poole and Rosenthal, 1985; Martin and Quinn, 2002) estimate the political ideologies of legisla- tors through their observed voting behavior, pos- sibly paired with the textual content of bills (Ger- rish and Blei, 2012) and debate text (Nguyen et al., 2015); other unsupervised models estimate ideolo- gies of politicians from their speeches alone (Sim et al., 2013). Twitter users have also been mod- eled in a similar framework, using their observed following behavior of political elites as evidence to be explained (Barber´a, 2015). Supervised mod- els, likewise, have not only been used for assessing the political stance of sentences (Iyyer et al., 2014) but are also very popular for predicting the holis- tic ideologies of everyday users on Twitter (Rao et al., 2010; Pennacchiotti and Popescu, 2011; Al Zamal et al., 2012; Cohen and Ruths, 2013; Volkova et al., 2014), Facebook (Bond and Mess- ing, 2015) and blogs (Jiang and Argamon, 2008), where training data is relatively easy to obtain— either from user self-declarations, political follow- ing behavior, or third-party categorizations. Aside from their intrinsic value, estimates of users’ political ideologies have been useful for quantifying the orientation of news media sources (Park et al., 2011; Zhou et al., 2011). We con- sider in this work a different task: estimating the political import of propositions like OBAMA IS A SOCIALIST. In focusing on propositional statements, we draw on a parallel, but largely independent, strand of research in open information extraction. IE sys- tems, from early slot-filling models with predeter- mined ontologies (Hobbs et al., 1993) to the large- scale open-vocabulary systems in use today (Fader et al., 2011; Mitchell et al., 2015) have worked toward learning type-level propositional informa- tion from text, such as BARACK OBAMA IS PRES- IDENT. To a large extent, the ability to learn these facts from text is dependent …

Content truncated to 500 words