Online hate speech has alarmingly increased as a result of social media platforms' explosive expansion, endangering people and communities. As a result, automatic identification of hateful text has emerged as a crucial area of study in machine learning and natural language processing. A multi-modal hate speech detection system that uses supervised machine learning approaches for efficient classification is presented in this paper. Naïve Bayes and Support Vector Machine (SVM) are two of the classification methods that the system uses to process textual input, utilising feature extraction techniques. A labelled dataset with both hateful and non-hateful content was used for the experimental evaluation. Confusion matrix analysis-based performance comparison reveals that the SVM model performs much better in terms of accuracy and decreased misclassification than the Naïve Bayes classifier. While reducing false positives, the suggested framework shows increased accuracy in identifying hateful content. The findings show that automated moderation and safer online communication environments can be successfully supported by machine learning-based multi-modal systems.
Mahesh et al. (Sun,) studied this question.