Sentiment Analysis on Twitter towards the Ratification of a Bill on the Elimination of Sexual Violence in Indonesia using Machine Learning

In Indonesia, incidents of violence against women have developed into a problem that needs attention. The National Commission for Women has recorded an increase in cases of violence during 2019 which was 6 percent compared to the previous year. The number of increases is a concern for the government in designing a countermeasure and prevention action. In 2014, the National Commission on Women initiated the A bill Law on the Elimination of Sexual Violence to follow up on acts of sexual violence in Indonesia. The ratification of a Bill on the Elimination of Sexual Violence is important to suppress cases of sexual violence. The pros and cons that arise regarding that are found in various media, including social media, namely Twitter as a forum for the community in freedom of expression to make a bill until now it has not been ratified. Comments in the form of tweets become a representation of sentiment that can be seen by the public, both positive and negative. This can affect a policy that is made whether it is feasible to apply or not. So that sentiment from the public can be analyzed properly, this study will examine sentiment analysis on Twitter regarding a Bill on the Elimination of Sexual Violence in Indonesia by applying Natural Language Processing (NLP), using the Support Vector Machine (SVM), and Naïve Bayes Classifier (NBC) algorithms with three different scenarios. The purpose of this study was to determine the algorithm with the best performance in classifying categories. From this research, the highest accuracy result for the test data is in scenario 3 with 97% using SVM and 94.50% using NBC. With these results, the model created can classify positive and negative categories in a document properly.