NLDS-UCSC at SemEval-2016 Task 6: A Semi-Supervised Approach to Detecting Stance in Tweets

Proceedings of SemEval-2016, pages 420–427, San Diego, California, June 16-17, 2016. c⃝2016 Association for Computational Linguistics NLDS-UCSC at SemEval-2016 Task 6: A Semi-Supervised Approach to Detecting Stance in Tweets Amita Misra, Brian Ecker, Theodore Handleman, Nicolas Hahn and Marilyn Walker Natural Language and Dialogue Systems Lab University of California, Santa Cruz 1156 N. High. SOE-3 Santa Cruz, California, 95064, USA amisra2|becker|thandlem|nhahn|mawalker@ucsc.edu Abstract Stance classification aims to identify, for a particular issue under discussion, whether the speaker or author of a conversational turn has Pro (Favor) or Con (Against) stance on the is- sue. Detecting stance in tweets is a new task proposed for SemEval-2016 Task6, involv- ing predicting stance for a dataset of tweets on the topics of abortion, atheism, climate change, feminism and Hillary Clinton. Given the small size of the dataset, our team created our own topic-specific training corpus by de- veloping a set of high precision hashtags for each topic that were used to query the twitter API, with the aim of developing a large train- ing corpus without additional human labeling of tweets for stance. The hashtags selected for each topic were predicted to be stance-bearing on their own. Experimental results demon- strate good performance for our features for opinion-target pairs based on generalizing de- pendency features using sentiment lexicons. 1 Introduction Social media websites such as microblogs, weblogs, and discussion forums are used by millions of users to express their opinions on almost everything from brands, celebrities, and events to important social and political issues. In recent years, the microblog- ging service Twitter has emerged as one of the most popular and useful sources of user content, and re- cent research has begun to develop tools and com- putational models for tweet-level opinion and senti- ment analysis. Stance classification aims to identify, for a particular issue under discussion, whether the speaker or author of a conversational turn has a Pro (Favor) or Con (Against) stance on the issue (So- masundaran and Wiebe, 2009; Somasundaran and Wiebe, 2010; Walker et al., 2012c; Sridhar et al., 2015; Hasan and Ng, 2013). Id Tweet Stance T1 We are causing the ice masses of Earth to melt at an alarming rate. FAVOR T2 ONE Volcano emits more pollution than man has in our HISTORY! AGAINST T3 It’s most exciting to witness a major develop- ment! NONE T4 The Weather app keeps taunting us with rain. #PNW #drought NONE Table 1: Example tweets with stance labels for the issue Climate Change is a Real Concern. Detecting stance in tweets is a new task proposed for SemEval-2016 (Mohammad et al., 2016). The aim of the task is to determine user stance (FAVOR, AGAINST, or NONE) in a dataset of tweets on the five selected topics of abortion, atheism, climate change, feminism and Hillary Clinton. Consider the tweets in Table 1, which express stance toward the target issue Climate Change is a Real Concern. It can be inferred that the author of tweet T1 is in favor of the target while the …

Content truncated to 500 words