Methods and Datasets for Detecting Hate Speech in Textual Content

Online discussions with harmful content can lead to group conflicts or abuse of online communities. Hate speech is complex, offensive, or harmful content targeting people or crowds. This paper detects textual hate speech, emphasizes the key datasets, text features, and machine learning models used, and systematically examines the deep learning technologies. This paper proposes a model for hate speech detection based on deep learning architecture. The proposed LSTM-CNN model to enhance the performance on benchmark datasets. This paper provides insights into the generic pipeline of automatic hate speech detection, including dataset collection, feature engineering, and model training. We compared the performance of our model with classical methods. It is prominent from the outcome that our proposed model performed with an accuracy of 93.1%.