EAF-HSD: Ensemble Adaptive Fuzzy Logic-Based Hate Speech Detection on Social Media

Abstract

With rise in incidence of hate speeches on social media platforms, this has become an important issue where sociability and human beings are among the greatest elements that are affected. Addressing this issue, an advanced algorithm for sentiment analysis and hate speech detection is proposed in this paper. This technique involves a pre-processing of social media data, which encompasses adaptive fuzzy logic-enhanced DBSCAN clustering for topic detection as well as semantic pattern recognition indicative of hate speech. An ensemble learning using Naive Bayes and Random Forest is used to detect the hate speech.The results confirm that the proposed EAF-HSD approach aids in reaching higher accuracy than the existing approaches. The accuracy of 94% shown in the model that has been proposed as far as classifying offensive language is concerned as the best of all and it has an overall accuracy of 92%, which shown a massive breakthrough over the baseline models. he proposed work introduces novel advancements in hate speech detection on social media platforms. Integrating adaptive fuzzy logic with adaptive DBSCAN clustering method to capture the intricate hate speech patterns. An ensemble learning framework combines diverse classifiers is used to classify the hate speech and non-hate speech accurately.