HASOCOne@FIRE-HASOC2020: Using BERT and Multilingual BERT models for Hate Speech Detection

TL;DR

This paper explores the use of BERT and multilingual BERT models for automatic hate speech detection, leveraging transfer learning on datasets from FIRE shared tasks, achieving high accuracy in classifying offensive content.

Contribution

It introduces the application of BERT-based models for hate speech detection on FIRE datasets, demonstrating their effectiveness over previous methods.

Findings

BERT and multilingual BERT models outperform previous approaches.

Transfer learning significantly improves classification accuracy.

The code is publicly available for reproducibility.

Abstract

Hateful and Toxic content has become a significant concern in today's world due to an exponential rise in social media. The increase in hate speech and harmful content motivated researchers to dedicate substantial efforts to the challenging direction of hateful content identification. In this task, we propose an approach to automatically classify hate speech and offensive content. We have used the datasets obtained from FIRE 2019 and 2020 shared tasks. We perform experiments by taking advantage of transfer learning models. We observed that the pre-trained BERT model and the multilingual-BERT model gave the best results. The code is made publically available at https://github.com/suman101112/hasoc-fire-2020.