Optimize Naïve Bayes Classifier Using Chi Square and Term Frequency Inverse Document Frequency For Amazon Review Sentiment Analysis

The rapid development of the internet has made information flow rapidly wich an impact on world commerce. Some people who have bought a product will write their opinion social media or other online site. Long-text buyer reviews need machine to recognize opinions. Sentiment analysis applies text mining method. One methods applied in sentiment is classification. classification algorithms naïve bayes classifier. Naïve classifier method with good efficiency and performance. However, it very sensitive too many features, makes accuracy low. To improve algorithm can be done by selecting features. feature selection chi square. features square calculation based top-K value that been determined, namely 450. In addition, weighting also algorithm. techniques term frequency inverse document (TF-IDF). this study, using labelled dataset (field amazon_labelled) obtained from UCI Machine Learning. This 500 positive negative reviews. amazon review was 82%. Meanwhile, applying TF-IDF 83%.