Hateful Text Detection in Social Media Using Machine Learning Algorithms

Abstract

The rise of social media has advanced unmatched access to communication facilities for human beings, which has been accompanied by a set of both positive and negative implications. One of the fastest and most rapidly growing challenges is the dissemination of hate speech, which threatens not only individuals but large societies too. This paper, therefore, undertakes an in-depth analysis of the machine learning algorithms that can be applied to hate text detection in social media platforms. A review of state-of-the-art approaches and datasets is carried out to review efficiencies by different models such as Support Vector Machines (SVM), Naive Bayes, deep learning algorithms like Convolutional Neural Networks, and Long Short-Term Memory Networks. These models are compared on the basis of various performance metrics such as accuracy, precision, recall, and F1-score. In this paper, we propose a hybrid model by fusing feature extraction methods such as TF-IDF, word embeddings, and deep neural networks to detect fake news more accurately. The results indicate that although each model was promising in their own ways, the proposed hybrid approach clearly improved the detection capability by reducing false positives and amplifying the overall performance.

DOI - https://doi.org/10.65525/SVUP.9788199963351.2026.133-140