A feature-adjusted naïve Bayes agorithm

Sentiment classification has been an important issue in natural language processing in recent years. In order to solve the distribution difference problem in cross-domain sentiment classification, we propose a featured-adjusted EM-based naïve Bayes algorithm which combines feature adaption and instance adaption simultaneously, and this algorithm can adjust the parameters in EM algorithm by the results of the feature adaption. The experimental results show that the proposed algorithm can improve the accuracy of cross-domain sentiment classification to some extent compared with the baseline algorithm.