Sentiment Analysis of Social Media Twitter with Case of Anti-LGBT Campaign in Indonesia using Naïve Bayes, Decision Tree, and Random Forest Algorithm

People use social media as a means to express their thoughts, interests, and opinions on various things. Thousands of submissions occur every day on every social media. Everyone can express their opinions through social media freely. These opinions contain positive, negative and neutral sentiments on a topic. The case study taken by researchers is the Anti-LGBT campaign in Indonesia. The case was taken because the Anti-LGBT campaign was widely discussed by the Indonesian people on Twitter’s social media. If you want to know the tendency of public comments on the Anti-LGBT campaign in Indonesia, is it positive, negative, or neutral, then a sentiment analysis is conducted. The algorithm used in conducting sentiment analysis is Naïve Bayes because it has a high degree of accuracy in classifying sentiment analysis. The stages in conducting sentiment analysis in this study are preprocessing data, processing data, classification, and evaluation. The sentiment analysis obtained in this study shows that Twitter users in Indonesia give more neutral comments. In this study, an accuracy of 86.43% was obtained from testing data using Naïve Bayes Algorithm in RapidMiner tools, where the accuracy is higher than the other algorithms, Decision Tree and Random Forest which is 82.91%.