Comparative Analysis of Text Classification Models for Offensive Language Detection on Social Media Platforms

Abstract

The detection of offensive language in text has
become increasingly crucial in various social media platforms
to maintain a respectful and safe environment. In this
research we study and present a comparative analysis of
different text classification models for identifying offensive
and non-offensive language. Specifically, we investigate the
performance of Support Vector Classifier (SVC), Compliment
model, Gaussian model, and Multinomial model on a dataset
curated for this purpose. Each text classification model is
implemented and trained using the preprocessed dataset, and
their performance is evaluated using standard evaluation
metrics such as accuracy. The experimental results display the
effectiveness of each model in distinguishing offensive
language from non-offensive language. This research
contributes to the literature by providing empirical evidence
on the performance of various text classification models for
offensive language detection, thus aiding in the development
of more robust and accurate detection systems for online
platforms.