9th International Conference on ICT for Sustainable Development

Authors - Preetam Vitthalkar, Prateek Deshpande, Vijeta Aarsanal, Satish chikkamath, Nirmala S R, Suneeta Budihal
Abstract - Sentiment analysis is the technique of analyzing people’s emotions, attitudes, and thoughts expressed in the form of written text. Detection and addressing of mental health issues early is vital for better outcomes and reducing overall societal costs. The majority of the research is based on opinion mining to comprehend the public’s opinion towards product perception. This work introduces a novel chatbot system to interact with users. The gathered user data is classified using a hybrid model combining Bi-LSTM and Na¨ıve Bayes classifier. The classifiers are trained on a concatenated dataset comprising reviews, tweets, comments, and texts, with additional comparison against various machine learning classifiers for accuracy assessment. Results indicate the effectiveness of the trained and tested model, achieving a high accuracy rate in predicting mental health conditions.