Citius: A Naive-Bayes Strategy for Sentiment Analysis on English Tweets

Proceedings of the 8th International Workshop on Semantic Evaluation (SemEval 2014), pages 171–175, Dublin, Ireland, August 23-24, 2014. Citius: A Naive-Bayes Strategy for Sentiment Analysis on English Tweets∗ Pablo Gamallo CITIUS Univ. de Santiago de Compostela pablo.gamallo@usc.es Marcos Garcia Cilenis Language Technology, S.L. marcos.garcia@cilenis.com Abstract This article describes a strategy based on a naive-bayes classifier for detecting the po- larity of English tweets. The experiments have shown that the best performance is achieved by using a binary classifier be- tween just two sharp polarity categories: positive and negative. In addition, in or- der to detect tweets with and without po- larity, the system makes use of a very basic rule that searchs for polarity words within the analysed tweets/texts. When the clas- sifier is provided with a polarity lexicon and multiwords it achieves 63% F-score. 1 Introduction Sentiment Analysis consists in finding the opin- ion (e.g. positive, negative, or neutral) from text documents such as movie reviews or product re- views. Opinions about movies, products, etc. can be found in web blogs, social networks, discus- sion forums, and so on. Companies can improve their products and services on the basis of the re- views and comments of their costumers. Recently, many works have stressed the microblogging ser- vice Twitter. As Twitter can be seen as a large source of short texts (tweets) containing user opin- ions, most of these works make sentiment analysis by identifying user attitudes and opinions toward a particular topic or product (Go et al., 2009). The task of making sentiment analysis from tweets is a hard challenge. On the one hand, as in any senti- ment analysis framework, we have to deal with hu- man subjectivity. Even humans often disagree on ∗This work has been supported by the projects: HPC- PLN: Ref:EM13/041 (Program Emergentes, Xunta de Gali- cia), Celtic: Ref:2012-CE138 and Plastic: Ref:2013-CE298 (Program Feder-Innterconecta) 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/ the categorization of the positive or negative sen- timent that is supposed to be expressed on a given text (Villena-Rom´an et al., 2013). On the other hand, tweets are too short text to be linguistically analyzed, and it makes the task of finding relevant information (e.g. opinions) much harder. The SemEval-2014 task “Sentiment Analysis in Twitter” is an evaluation competition that in- cludes a specific task directly related to sentiment analyisis. In particular, subtask B, called “Mes- sage Polarity Classification”, consists in classify- ing whether a given message is of positive, neg- ative, or neutral sentiment. For messages con- veying both a positive and negative sentiment, the stronger sentiment should be chosen. The results of our system in this task are situated in the aver- age out of 51 evaluated systems. In this article, we describe the learning strate- gies we developed so as to perform this task, all of them based on bayesian classification. 2 Naive Bayes Classifier Most of the algorithms for …

Content truncated to 500 words