WikiSent: weakly supervised sentiment analysis through extractive summarization with wikipedia

摘要: This paper describes a weakly supervised system for sentiment analysis in the movie review domain. The objective is to classify into polarity class, positive or negative, based on those sentences bearing opinion alone, leaving out other irrelevant text. Wikipedia incorporates world knowledge of movie-specific features which used obtain an extractive summary review, consisting reviewer's opinions about specific aspects movie. filters concepts are with respect given proposed system, WikiSent, does not require any labeled data training. It achieves better comparable accuracy existing semi-supervised and unsupervised systems domain, same dataset. We also perform general trend using WikiSent.