• Version
• Download 7
• File Size 338.39 KB
• File Count 1
• Create Date 19 May 2023
• Last Updated 19 May 2023

TOXIC CONTENT CLASSIFICATION USING MACHINE LEARNING

Ashika Ruth Saldana, Athmi B.S, Harshitha H.S, Shreya
A J Institute of Engineering and Technology, Mangalore

ABSTRACT
Online toxic content has become a global issue as the number of internet users is growing and technologies are advancing. The cyber-world is a space for everyone, irrespective of their educational and cultural backgrounds. Identifying and differentiating hate content from other toxic content in the cyber environment is a challenging task for automated systems. Classifying toxic content is a difficult task because it involves text processing and context understanding. Social networking platforms have grown in popularity and are used for a variety of activities such as product promotion, news sharing, and achievement sharing, among others. On the other hand, it is also used to spread rumors, bully people, and target specific groups of people. Hate and offensive posts must be detected and removed from social media platforms as soon as possible because they spread quickly and have a wide range of negative consequences for people. In recent years, offensive content and toxic content detection have become popular research topics. Toxic Content classification is an approach to automatically classify toxic content on Twitter into two classes: hate content and non-hate content. We use different features such as a bag of words, term frequency-inverse document frequency (TFIDF), and N-Grams to train and test machine learning algorithms. We also perform a comparative analysis of the different machine learning models. Classifying toxic content is a difficult task because it involves text processing and context understanding. In our approach, we aim to demonstrate a significant improvement in toxic content classification.

Keywords:

Hate Speech, Machine Learning, ANN, SVM, KNN, Naive Bayes, Twitter