Analysis Sentiment of Public Opinion on Social Media Using Naïve Bayes and TF-IDF Algorithms

This research examines the performance of the Naive Bayes and TF-IDF algorithms in carrying out sentiment analysis on social media. When using applications, people need reviews from other users about opinions, feelings, judgments, attitudes and emotions based on language. Sentiment analysis can be used in this case to review public opinion. One of the methodologies created by the emergence of this problem is classification to analyze positive or negative sentiment. This classification method is used to see the results of sentiment analysis of review data on social media using the Naive Bayes algorithm and TF-IDF. By using the implementation of the classification method on public opinion sentiment, this research found that from 4150 data that had been analyzed, the ratio of positive sentiment was 2479 and negative sentiment was 1490 data.