Sentiment analysis is a common task in natural language processing that aims detect polarity of a text document (typically a consumer review). In the settings, we discriminate only between positive and negative, turning the task into a standard binary classification problem. We several ma- chine learning approaches to this problem, and combine them achieve the best possible results. We show how to use for this task the generative lan- guage models, which are slightly complementary to the of the art techniques. We achieve strong results on a well-known dataset IMDB movie reviews. Our results are easily reproducible, as we publish also code needed to repeat the experiments. This should simplify further advance the state of the art, as other researchers can combine their techniques with with little effort.
No takes yet. Share an insight, caveat, or question.
Mesnil et al. (2014) studied this question.