Basic Structure for Sentiment and Emotion Analysis-Based Hate Speech Recognition and Integrating of Unstructured Data on Facebook, or Fadohs

Any remarks spoken disparagingly about an individual or group due to their sexual orientation, religion, race, or any other attribute constitute hate speech. Even though there are other offline distribution methods, the growing popularity of social networking sites has significantly increased its use and seriousness. Thus, the main objective of this research is to find and examine the disorganized data of particular social media posts in order to spread hate speech in the comment sections. Our suggested FADOHS architecture addresses this problem by warning all social media companies about the ubiquity of hate speech on these platforms through the combination of data assessment and natural language processing methods. Specifically, we use belief and emotion assessment algorithms to analyze these websites and contemporary interactions. Refined messages that might contain dehumanizing phrases will be sent to the clustering algorithm first. Preliminary findings indicate that the proposed FADOHS framework can achieve approximately 10 % improvement in accuracy, recall, and F1 scores over the state-of-the-art approach.