Hate Speech Detection using ML algorithms

Abstract

Social media is a growing platform where different users share their ideas and sentiments towards different topics because users spend a lot of time expressing their thoughts and views. There are various researches going on in detecting the sentiments of the user’s comments but the main sentiment factor remain undiagnosed. In this paper, the aim is to detect hate speeches. The dataset was preprocessed and cleaned and cleaned text was explored to get a better understanding. Salient features were extracted from the data to train our model and to identify the hate sentiments of tweets. The vector model is created using genism to learn the relationship between words and based on that sentence are labeled. Stop words and port stemmer are used to filter unwanted data to build the vocabulary using CountVectorizer before it is used for model building. Using various machine algorithms, comparative study is done to check the performance of algorithms and promising results are attained.