Sentiment Analysis using Naive Bayes in Rude Word Replacement on Digital Platforms

Abstract

Indonesia has entered the era of revolution 4.0, which in fact uses a lot of digital access in its daily life. One of the most popular digital accesses is social media, a place where everyone can connect with each other online, which certainly brings good and bad impacts. An example of a bad impact that we often encounter is the use of harsh words to fellow social media users. This research aims to minimize the existence of the use of harsh words by filtering every harsh word that appears. The research uses Machine Learning with the Naïve Bayes method by training a dataset of harsh words and a special dictionary for subtle words as a translator in a more subtle language. The results of the research are expected to provide insight into the accuracy of using the Naïve Bayes method and provide output of word filtering features that have an impact on the character building of digital platform users anywhere and anytime.