Réseau de neurones profond et SVM pour la classification des sentiments

As web 2.0 is spreading, users get used to give there opinions on forums, blogs and e-commerce websites. This is a valuable piece of information for many applications such as consumer modeling, sales prediction or opinion survey. According to the literature, the efficiency of opinion mining tools will mainly relies on the ability of discriminating texts that express positive sentiments from texts that express negative ones. Previous experiments show that this task is difficult. Our approach relies on deep neural network (NN), mixing standard layers and convolutionnal layers. We demonstrate the efficiency of our approach on classical Amazon reviews data-set: raw performance equals Support Vector Machine (SVM) and the state of the art but our predictions differ significantly from SVM and combining SVM and NN enables us to overcome SVM. MOTS-CLÉS : Classification de sentiments, réseaux de neurones profonds