Does Twitter Sentiment Movie Review Using Naive Bayes and Feature Selection Information Gain Effective?

The opinion of Twitter users related with movie is a valuable asset to be studied. This is because business and organization casts always want to know the public and consumers' opinions about their movies and consumers also want to know others' opinion before purchasing movie products. Sentiment analysis was applied in this research to gain information whether a tweet is positive opinion, negative opinion, or neutral opinion. 

First of all, i don't really think using Naive Bayes Algorithm for manage Twitter review is effective , although because it can reach the highest value of accuration rather than another statistical algorithm. One problem in the sentiment classification is if there are too many attributes used in the dataset then the accuracy getting lower . The Nave Bayes Classifier method it self also have the same weakness which is sensitivity to the too many attributes or features that it reduces the accuracy level. In my opinion the weakness can be overcome by using feature selection algorithm. 

Moreover, it can increase the accuracy level and the time efficiency of the classifier. Some research has done with this problem. Chandani and Wahono in 2015 compared five most appropriate feature selection algorithms for the sentiment analysis on the movie review dataset and the information gain algorithm gained the highest average classification accuracy with percentage of 84.57%, also Sihwi, Jati & Anggrainingsih in 2018 successfully gets the 82.19% accuracy level using this combination method, which is if the accuracy level up to 80% it can be assume that the method's performance is excellent.

From all of it, i believe that Twitter Sentiment Movie Review Using Naive Bayes and Feature Selection Information Gain is effective. I think the weakness of Naive Bayes algorithm can be covered by Information Gain, and the accuracy can reach the high level.