Negation-Aware Framework for Sentiment Analysis in Arabic Reviews

Abstract

Sentiment analysis deals with identifying polarity orientation embedded in users' comments and reviews. It aims at discriminating positive reviews from negative ones. Sentiment is related to culture and language morphology. In this paper, we investigate the effects of language morphology on sentiment analysis in reviews written in the Arabic language. In particular, we investigate, in details, how negation affects sentiments. We also define a set of rules that capture the morphology of negations in Arabic. These rules are then used to detect sentiment taking care of negated words. Experimentations prove that our suggested approach is superior to several existing methods that deal with sentiment detection in Arabic reviews.