COMMENT TOXICATION USING DEEP LEARNING NETWORKS

Online communication has been plagued by the issue of toxic comments more and more. The goal of this research is to identify how toxic a specific comment is. Comments using explicit language can be categorized as Toxic, Severe Toxic, Obscene, Threatening, Insulting, and Identity Hating. To identify and categorise hazardous remarks, the effort uses natural language processing techniques and machine learning algorithms. The project’s objective is to create a comprehensive model that can be used to recognise and censor harmful remarks in real time across a range of online venues. To preprocess the text data, extract features, and construct the classification model, the project employs a mix of natural language processing (NLP) techniques and machine learning (ML) algorithms. Tokenization is used in the preprocessing steps, and word embeddings are used in the feature extraction phase.We pass this preprocessed data through the deep learning model. Next, we’ll assess the models’ performance before finishing by building a Gradio app. The project resulted in a model capable of identifying various kinds of toxic comments . The model can be integrated into online platforms to allow automatic comment moderation and to promote healthy online discussions. The project has far-reaching consequences for promoting online safety, reducing cyberbullying, and cultivating a positive online environment.