Implicit and Explicit Sarcasm Detection via Target of Offensive Language and Cognition of Logistic Tsallis Entropy Distribution in Social Networks

Social media platforms have become increasingly popular avenues for individuals to express their thoughts and emotions, especially with the advancements in computer systems and other technologies. One area of research that holds significant commercial potential is the collection and analysis of sentiments expressed in user comments, particularly in terms of their polarities. The detection, identification, and interpretation of sarcasm are of particular importance in sentiment analysis. However, automated sarcasm detection is a complex and challenging task that has not been extensively explored. This paper presents a novel approach for detecting sarcasm in social media, specifically on platforms such as Twitter and Instagram, by utilizing unsupervised mathematical optimization and probability distribution techniques. In this study, we initially employ a probabilistic distribution categorization method and a lexicon-based framework to analyze the words. Subsequently, the VADER sentiment library is utilized to ascertain the polarities of the specified sentences. By incorporating a parameter specifically designed to detect sarcasm in sentences, a sentiment score is then fed into the logistical distribution. Ultimately, we present statements that exhibit both implicit and explicit sarcasm, as well as those that contain offensive language. This proposed model serves to safeguard social media platforms against the dissemination of sarcasm, the propagation of abusive content, and instances of cyberbullying.