Detecting Stance in Tweets And Analyzing its Interaction with Sentiment

Proceedings of the Fifth Joint Conference on Lexical and Computational Semantics (*SEM 2016), pages 159–169, Berlin, Germany, August 11-12, 2016. Detecting Stance in Tweets And Analyzing its Interaction with Sentiment Parinaz Sobhani EECS, University of Ottawa psobh090@uottawa.ca Saif M. Mohammad and Svetlana Kiritchenko National Research Council Canada {saif.mohammad,svetlana.kiritchenko}@nrc-cnrc.gc.ca Abstract One may express favor (or disfavor) to- wards a target by using positive or neg- ative language. Here for the first time we present a dataset of tweets annotated for whether the tweeter is in favor of or against pre-chosen targets, as well as for sentiment. These targets may or may not be referred to in the tweets, and they may or may not be the target of opin- ion in the tweets. We develop a sim- ple stance detection system that outper- forms all 19 teams that participated in a re- cent shared task competition on the same dataset (SemEval-2016 Task #6). Addi- tionally, access to both stance and senti- ment annotations allows us to conduct sev- eral experiments to tease out their interac- tions. We show that while sentiment fea- tures are useful for stance classification, they alone are not sufficient. We also show the impacts of various features on detect- ing stance and sentiment, respectively. 1 Introduction Stance detection is the task of automatically de- termining from text whether the author of the text is in favor of, against, or neutral towards a propo- sition or target. The target may be a person, an organization, a government policy, a movement, a product, etc. For example, one can infer from Barack Obama’s speeches that he is in favor of stricter gun laws in the US. Similarly, people of- ten express stance towards various target entities through posts on online forums, blogs, Twitter, Youtube, Instagram, etc. Automatically detecting stance has widespread applications in information retrieval, text summa- rization, and textual entailment. Over the last decade, there has been active research in model- ing stance. However, most work focuses on con- gressional debates (Thomas et al., 2006) or de- bates in online forums (Somasundaran and Wiebe, 2009; Murakami and Raymond, 2010; Anand et al., 2011; Walker et al., 2012; Hasan and Ng, 2013). Here we explore the task of detecting stance in Twitter—a popular microblogging plat- form where people often express stance implicitly or explicitly. The task we explore is formulated as follows: given a tweet text and a target entity (person, orga- nization, movement, policy, etc.), automatic natu- ral language systems must determine whether the tweeter is in favor of the given target, against the given target, or whether neither inference is likely. For example, consider the target–tweet pair: Target: legalization of abortion Tweet: The pregnant are more than walking(1) incubators, and have rights! Humans can deduce from the tweet that the tweeter is likely in favor of the target.1 Note that lack of evidence for ‘favor’ or ‘against’, does not imply that the tweeter is neu- tral towards the target. It may just mean that we cannot deduce stance …

Content truncated to 500 words