LT3: Sentiment Classification in User-Generated Content Using a Rich Feature Set

Proceedings of the 8th International Workshop on Semantic Evaluation (SemEval 2014), pages 406–410, Dublin, Ireland, August 23-24, 2014. LT3: Sentiment Classification in User-Generated Content Using a Rich Feature Set Cynthia Van Hee, Marjan Van de Kauter, Orph´ee De Clercq, Els Lefever and V´eronique Hoste LT3, Language and Translation Technology Team Department of Translation, Interpreting and Communication – Ghent University Groot-Brittanni¨elaan 45, 9000 Ghent, Belgium Firstname.Lastname@UGent.be Abstract This paper describes our contribution to the SemEval-2014 Task 9 on sentiment analysis in Twitter. We participated in both strands of the task, viz. classification at message-level (subtask B), and polarity disambiguation of particular text spans within a message (subtask A). Our experi- ments with a variety of lexical and syntactic fea- tures show that our systems benefit from rich fea- ture sets for sentiment analysis on user-generated content. Our systems ranked ninth among 27 and sixteenth among 50 submissions for task A and B respectively. 1 Introduction Over the past few years, Web 2.0 applications such as microblogging services, social network- ing sites, and short messaging services have con- siderably increased the amount of user-generated content produced online. Millions of people rely on these services to send messages, share their views or gather information about others. Si- multaneously, companies, marketeers and politi- cians are anxious to detect sentiment in UGC since these messages might contain valuable informa- tion about the public opinion. This explains why sentiment analysis has been a research area of great interest in the last few years (Wiebe et al., 2005; Wilson et al., 2005; Pang and Lee, 2008; Mohammad and Yang, 2011). Though first studies focussed more on product or movie reviews, we see that analyzing sentiment in UGC is currently becoming increasingly popular. The main differ- ence between these two sources of information is that the former is rather long and contains quite formal language whereas the latter one is gener- ally very brief and noisy and thus represents some different challenges (Maynard et al., 2012). In this paper, we describe our contribution to the SemEval-2014 Task 9: Sentiment Analysis in This work is licensed under a Creative Commons At- tribution 4.0 International Licence. Page numbers and pro- ceedings footer are added by the organisers. Licence details: http://creativecommons.org/licenses/by/4.0/ Twitter (Rosenthal et al., 2014), which was a rerun of SemEval-2013 Task 2 (Nakov et al., 2013) and consisted of two subtasks: • Subtask A - Contextual Polarity Disambiguation: Given a message contain- ing a marked instance of a word or phrase, determine whether that instance is positive, negative or neutral in that context. • Subtask B - Message Polarity Classification: Given a message, classify whether the message is of positive, negative, or neutral sentiment. For messages convey- ing both a positive and negative sentiment, whichever is the stronger sentiment should be chosen. The datasets for training, development and test- ing were provided by the task organizers. The training datasets consisted of Twitter messages on a variety of topics. The test sets con- tained regular tweets (Twitter2013, Twitter2014), tweets labeled …

Content truncated to 500 words