SSA-UO: Unsupervised Sentiment Analysis in Twitter

This paper describes the specifications and results of SSA-UO, unsupervised system, presented in SemEval 2013 for Sentiment Analysis in Twitter (Task 2) (Wilson et al., 2013). The proposal system includes three phases: data preprocessing, contextual word polarity detection and message classification. The preprocessing phase comprises treatment of emoticon, slang terms, lemmatization and POS-tagging. Word polarity detection is carried out taking into account the sentiment associated with the context in which it appears. For this, we use a new contextual sentiment classification method based on coarse-grained word sense disambiguation, using WordNet (Miller, 1995) and a coarse-grained sense inventory (sentiment inventory) built up from SentiWordNet (Baccianella et al., 2010). Finally, the overall sentiment is determined using a rule-based classifier. As it may be observed, the results obtained for Twitter and SMS sentiment classification are good considering that our proposal is unsupervised.