Acquiring an Italian polarity lexicon through distributional methods

Recent interests in Sentiment Analysis brought the attention on effective methods to detect opinions and sentiments in texts. Many approaches in literature are based on resources, such as Polarity Lexicons, which model the prior polarity of words or multi-word expressions. Developing such resources is expensive, language dependent, and linguistic sentiment phenomena are not fully covered in them. In this paper an automatic method for deriving polarity lexicons based on Distributional Models of Lexical Semantics is presented. Given a set of heuristically annotated messages from Twitter, we transfer sentiment information from sentences to words. As the approach is mostly unsupervised, it enables the acquisition of polarity lexicons for languages that are lacking these resources.We acquired a polarity lexicon in the Italian language, and experiments on Sentiment Analysis tasks show the benefit of the generated resources.