Sentiment analysis plays a pivotal role in comprehending user emotions within the Social Web; however, accurately capturing nuanced sentiment remains challenging. Although deep learning models exhibit impressive performance, their lack of transparency and interpretability hinders their real-world applicability. To address this limitation, we propose an Updated SHAP (SHapley Additive exPlanations) method integrated with Explainable Artificial Intelligence (XAI) as a solution. The research investigates the potential of XAI in enhancing sentiment analysis by designing, evaluating, and utilizing XAI models. The primary focus is on knowledge discovery, exploring how XAI aids in recognizing, interpreting, and simulating human emotions for sentiment analysis tasks. By illuminating the inner workings of sentiment analysis models, this research aims to substantially improve the reliability and utility of sentiment analysis in socio-affective domains. Emphasizing transparency and trust, XAI fosters informed decision- making and enhances sentiment analysis applications amidst the intricate landscape of user emotions on the Social Web.