Sentiment Analysis of Social Media Users Using Naïve Bayes, Decision Tree, Random Forest Algorithm: A Case Study of Draft Law on the Elimination of Sexual Violence (RUU PKS)

In this digitalization era, it made it easy for the people of Indonesia to deliver ideas, opinions to carry out campaigns. One real example is regarding the existence of Draft Law on the Elimination of Sexual Violence (RUU PKS) with the aim of creating a new paradigm that ensures the community is free from sexual violence. The draft initiated by the National Commission on Violence Against Women, known as KOMNAS Perempuan, succeeded in becoming a conversation with social media users. Twitter is one of the social media that is a means of delivering public opinion regarding the case. One of the uses of this research is to find out the tendency of Twitter users to comment on regarding the existence of Draft Law on the Elimination of Sexual Violence (RUU PKS) by conducting sentiment analysis. The stages of this research are carried out by collecting data. The data that has been collected will go through the pre-processing, processing stage which will produce data in the positive, negative and neutral categories then testing using confession matrix. From the results of accuracy in the order of naive bayes, decision tree and random forest, the accuracy is 83.94%, 75.31 and 75.72%. Based on this, it can be concluded that the sentiment analysis of this case gets the best results using the Naive Bayes algorithm. As well as based on the results obtained that campaigning through social media has an effect, in this case for the campaign to support the RUU PKS itself produces more positive sentiments.