In this paper we consider a new approach for domain-specific sentiment lexicon extraction in Russian. We propose a set of statistical features and algorithm combination that can discriminate sentiment words in a specific domain. The extraction model is trained in the movie domain and then utilized to other domains. We evaluate the quality of obtained sentiment vocabularies intrinsically. Finally we combine the sentiment lexicons from five domains to obtain one general lexicon for the product

more »

... a-domain. We demonstrate the robustness of the extracted lexicon in the cross-domain sentiment classification in Russian.