Comment Toxicity Model Using BiLSTM

The rise of social and communication platforms has led to an explosion of user-generated content. However, this growth has also triggered an increase in toxic and offensive remarks in both text and audio formats. Such harmful comments contribute to a hostile online environment and can negatively impact individuals and communities. To mitigate this issue, developing automated tools for detecting toxic content has become imperative. This paper introduces a toxic comment analyzer powered by BiLSTM, showcasing the effectiveness of pre-trained language models in identifying toxic language, with an achieved accuracy of 90% on datasets.