Analysis and Prediction of the Sentiments of the WhatsApp Chat Using Sentiment Analysis

Sentiment analysis is invaluable for understanding public opinion, especially for social media monitoring, feedback analysis and marketing strategies. With the rise of social media platforms like WhatsApp, billions of users share their thoughts and feelings every day, especially in group chats. This study dives into the fascinating field of how people experience emotions during communication and provides a comprehensive solution through emotion analysis. The “WhatsApp Chat Analyser” web application adapted for this purpose allows you to analyse WhatsApp group chats. Using Python libraries such as NLTK, Matplotlib, Re, Seaborn, Streamlit and Panda, we present a new approach to distinguish individual thoughts and feelings in a valence, arousal and dominance (VAD) framework and classify WhatsApp group messages. WhatsApp text conversations often evoke curiosity, arousal and a sense of control over emotional dimensions within VAD. A standard support vector machine (SVM) model and language classification are used to assess mood, with mood ratings classified based on the VAD scale. Qualitative content analysis forms the core of our data analysis. The results highlight the central role of WhatsApp groups in sharing information, exchanging ideas and discussing issues, with neutral or positive feelings about the chosen topics. WhatsApp chat data is pre-processed, sentiment analysis of individual messages are normalised using the recommended approach to obtain an overall score. This study improves our understanding of emotional dynamics in WhatsApp group chats and highlights the crucial role of the NLTK library in facilitating linguistic analysis of sentiment evaluation.