Feature selection for opinion polarity detection by machine learning method

Abstract

Recent years have seen increasing interest in techniques of opinion mining andsubjectivity analysis. In this article, we outline the results generated by our approach todetecting features for the classification polarity of opinions in French language using machinelearning techniques. Indeed, in sentiment analysis, identifying features associated with anopinion can help to produce a finer-grained understanding of subjective previews. In thisarticle, the proposed system consists of three phases: the pretreatments of the corpus, theextraction of the features and the classification. The second phase of our work represents thecombination of the co-occurrence analysis for a better management of the intrinsic semanticsof a word carrying opinion, and therefore a better extraction of features for classification.