Linguistically-motivated and Lexicon Features for Sentiment Analysis of Italian Tweets

In this paper we describe our approach to EVALITA 2014 SENTIment POLarity Classification (SENTIPOLC) task. We participated only in the Polarity Classification sub-task. By resorting to a wide set of general-purpose features qualifying the lexical and grammatical structure of a text, automatically created ad-hoc lexicons and existing free available resources, we achieved the second best accuracy.