Why are You Taking this Stance? Identifying and Classifying Reasons in Ideological Debates

Proceedings of the 2014 Conference on Empirical Methods in Natural Language Processing (EMNLP), pages 751–762, October 25-29, 2014, Doha, Qatar. c⃝2014 Association for Computational Linguistics Why are You Taking this Stance? Identifying and Classifying Reasons in Ideological Debates Kazi Saidul Hasan and Vincent Ng Human Language Technology Research Institute University of Texas at Dallas Richardson, TX 75083-0688 {saidul,vince}@hlt.utdallas.edu Abstract Recent years have seen a surge of interest in stance classification in online debates. Oftentimes, however, it is important to de- termine not only the stance expressed by an author in her debate posts, but also the reasons behind her supporting or oppos- ing the issue under debate. We therefore examine the new task of reason classifi- cation in this paper. Given the close in- terplay between stance classification and reason classification, we design computa- tional models for examining how automat- ically computed stance information can be profitably exploited for reason classifica- tion. Experiments on our reason-annotated corpus of ideological debate posts from four domains demonstrate that sophisti- cated models of stances and reasons can indeed yield more accurate reason and stance classification results than their sim- pler counterparts. 1 Introduction In recent years, researchers have begun exploring new opinion mining tasks. One such task is debate stance classification (SC): given a post written for a two-sided topic discussed in an online debate fo- rum, determine which of the two sides (i.e., for or against) its author is taking (Agrawal et al., 2003; Thomas et al., 2006; Bansal et al., 2008; Soma- sundaran and Wiebe, 2009; Burfoot et al., 2011; Hasan and Ng, 2013b). For example, the author of the post shown in Figure 1 is pro-abortion. Oftentimes, however, it is important to deter- mine not only the author’s stance expressed in her debate posts, but also the reasons why she supports or opposes the issue under debate. Intuitively, given a debate topic such as “Should abortion be banned?” or “Do you support Obamacare?”, it [I feel that abortion should remain legal, or rather, parents should have the power to make the decision themselves and not face any legal hindrance of any form.]1 Let us take a look from the social perspective. [If parents cannot afford to provide for the child, or if the family is facing financial constraints, it is understandable that abortion can remain as one of the options.]2 Reason 1: Woman’s right to abort Reason 2: Unwanted babies are threat to their parents’ fu- ture Figure 1: A sample post on abortion annotated with reasons. should not be difficult for us to come up with a set of reasons people typically use to back up their stances. Given a set of reasons associated with each stance in an online debate, the goal of post- level reason classification is to identify those rea- son(s) an author uses to back up her stance in her debate post. A more challenging version of this task is sentence-level reason classification, where the goal is to identify not only the reason(s) an au- …

Content truncated to 500 words