Comparison of the TF-IDF Method with the Count Vectorizer to Classify Hate Speech

Abstract

Hate speech is a form of expression used to spread hatred and commit acts of violence and discrimination against a person or group of people for various reasons. Cases of hate speech are very common in social media, one of which is Twitter. The goal to be achieved is to create a system that can classify a tweet on Twitter into hate speech (HS) or non-hate speech (NONHS) classes. The method used is Support Vector Machine by comparing the features of TF-IDF and Count Vectorizer. And the parameters compared are seen from accuracy, precision, recall, and f1-score. Results obtained, overall, by using the TF-IDF feature, the Support Vector Machine algorithm gets high results compared to the Count Vectorizer feature, with an accuracy value of 88.77%, 87.45% precision, 88.77% recall, and f1-score of 87.81%.