Toxic Comment Classification and Mitigation in Social Media Platforms

Sunitha Sabbu, Rajani D., Shaik Jaheed D., Sreeja Y., Venkata Naga Hemanth Reddy B.

2025

Abstract

Toxic online discussions have become a growing concern on digital platforms, necessitating automated solutions for detecting harmful content. This paper presents a Deep learning-based system that uses NLP methods to classify comments. Comments is categorized into toxic or non-toxic-are recognized by the system. A LSTM neural network-based classifier, integrated with Word2Vec skip-gram embedding and a fully connected network, is employed to process and classify user-generated comments. To facilitate real-time toxicity detection, we integrated the trained model with Twitter clone that allows users to input comments and receive toxicity predictions instantly. The dataset, sourced from a publicly available toxic comment corpus, consists of over 110,480 comments, providing a robust foundation for training. Model evaluation demonstrates high accuracy and Precision, with misclassification challenges observed in sarcasm and implicit toxicity detection. This research contributes to the field of online content moderation, offering an efficient and scalable approach for classifying comments. Future enhancements include advanced deep learning models and embedded context to improve nuanced toxicity identification in internet-based discussions.

Paper Citation

in Harvard Style

Sabbu S., D. R., D. S., Y. S. and B. V. (2025). Toxic Comment Classification and Mitigation in Social Media Platforms. In Proceedings of the 1st International Conference on Research and Development in Information, Communication, and Computing Technologies - ICRDICCT`25; ISBN 978-989-758-777-1, SciTePress, pages 627-631. DOI: 10.5220/0013903100004919

in Bibtex Style

@conference{icrdicct`2525,
author={Sunitha Sabbu and Rajani D. and Shaik D. and Sreeja Y. and Venkata B.},
title={Toxic Comment Classification and Mitigation in Social Media Platforms},
booktitle={Proceedings of the 1st International Conference on Research and Development in Information, Communication, and Computing Technologies - ICRDICCT`25},
year={2025},
pages={627-631},
publisher={SciTePress},
organization={INSTICC},
doi={10.5220/0013903100004919},
isbn={978-989-758-777-1},
}

in EndNote Style

TY - CONF

JO - Proceedings of the 1st International Conference on Research and Development in Information, Communication, and Computing Technologies - ICRDICCT`25
TI - Toxic Comment Classification and Mitigation in Social Media Platforms
SN - 978-989-758-777-1
AU - Sabbu S.
AU - D. R.
AU - D. S.
AU - Y. S.
AU - B. V.
PY - 2025
SP - 627
EP - 631
DO - 10.5220/0013903100004919
PB - SciTePress