Comparative analysis of sentiment orientation using SVM and Naive Bayes techniques

In the recent few years several efforts were dedicated for mining opinions and sentiment automatically from natural language in online networking messages, news and business product reviews. In this paper, we have explored sentiment orientation considering the positive and negative sentiments using film user reviews. We applied the technique Naive Bayes' classifier.). We have performed the sentiment analysis on the reviews using the algorithms like Naive Bayes, Linear SVM and Synthetic words. Our experimental results indicate that the Linear SVM has provided the best accuracy which is followed by the Synthetic words approach. The result also evaluate that the highest accuracy rate is of drama.