Dataset: Bengali Online Comments

This dataset contains almost 16,000 Bengali comments collected manually, each annotated with labels indicating different categories of content: troll, sexual, religious, threat, and non-bully. The data is organized in an Excel file. Designed for NLP tasks such as content classification, moderation, and toxicity detection, this dataset is particularly useful for applications in social media monitoring, automated content moderation, and the analysis of user-generated content in Bengali, enabling the detection and filtering of potentially harmful or inappropriate comments.