Assessing the Level of Toxicity Against Distinct Groups in Bangla Social Media Comments: A Comprehensive Investigation

TL;DR

This study develops a dataset and employs transformer models to detect and categorize toxic comments in Bangla social media posts targeting specific groups, revealing model effectiveness and toxicity complexities.

Contribution

It introduces a new dataset and applies transformer-based models to classify toxicity levels in Bangla social media comments targeting marginalized groups.

Findings

Bangla-BERT achieves an F1-score of 0.8903.

Toxic comments vary significantly across different demographic groups.

Transformer models effectively identify toxicity in Bangla social media texts.

Abstract

Social media platforms have a vital role in the modern world, serving as conduits for communication, the exchange of ideas, and the establishment of networks. However, the misuse of these platforms through toxic comments, which can range from offensive remarks to hate speech, is a concerning issue. This study focuses on identifying toxic comments in the Bengali language targeting three specific groups: transgender people, indigenous people, and migrant people, from multiple social media sources. The study delves into the intricate process of identifying and categorizing toxic language while considering the varying degrees of toxicity: high, medium, and low. The methodology involves creating a dataset, manual annotation, and employing pre-trained transformer models like Bangla-BERT, bangla-bert-base, distil-BERT, and Bert-base-multilingual-cased for classification. Diverse assessment metrics such as accuracy, recall, precision, and F1-score are employed to evaluate the model's effectiveness. The experimental findings reveal that Bangla-BERT surpasses alternative models, achieving an F1-score of 0.8903. This research exposes the complexity of toxicity in Bangla social media dialogues, revealing its differing impacts on diverse demographic groups.