Application of The Naïve Bayes Algorithm for Hate Speech Detection

Abstract

The spread of hate speech on social media using Indonesian continues to increase. Hate speech has become a problem that is detrimental to human life in society. In this research, the Naive Bayes algorithm is implemented to detect hate speech in social media comments. The dataset used in this research consists of 10,000 Indonesian language social media comments labeled as “Hate Speech” and “Non-Hate Speech”. Accuracy for data processed using TF-IDF is 85% for Indonesian language data. Based on the results, Naive Bayes is a simple but effective analytical approach in detecting hate speech.