Deep Learning for Toxic Comment Detection in Online Platforms

Abstract

This research aimed to explore the use of Deep Learning Artificial Neural Networks (ANNs) for toxic comment classification on social media and online forums. The prevalence of toxic interactions on these platforms has reached an all-time high, resulting in a decline in digital civility. The research reviews various algorithms and techniques for building promising ANNs and compares the performance of three chosen models on the Kaggle competition dataset to determine the best-performing ANN for this data problem.

The study includes a detailed overview of the background and techniques used for building and achieving the targeted results using Python and its libraries. It also covers technical aspects such as the building process, training, results, and evaluation, comparing the datasets using graphs and similar methods with the help of Python and its libraries. Classifying the nature of hate comments will provide flexibility for platforms to deal with them and open doors for new discussions and solutions.