A Deep Learning-Based Framework for Opinion Mining of Patient Reviews on Web Forums

The number of users and daily tweets on Twitter is among the highest of all social media outlets. New forms of communication like social media greatly contribute to the rapid treatment of panic amid global health crises like pandemics. The purpose of this research is to examine and evaluate the tone of the Tweets that were posted during the MONKEYPOX disease, for that purpose Long Short-Term Memory (LSTM), CNN, and a hybrid model of CNN-LSTM is used, in this the negative and positive sentiment classes of tweets are found. The findings indicate that there is only a tiny difference between positive and negative tweets, which highlights the necessity for attention to be paid to increase awareness and, as a consequence, decrease negativity. In addition, the LSTM was able to accurately predict the sentiment classes (Negative and Positive), with an accuracy rate of 72.15%. CNN with an accuracy of 82% and hybrid model CNN-LSTM with an accuracy of 91% The great accuracy leads one to the conclusion that Hybrid model CNN-LSTM is capable of adapting to new situations and predicting the emotion of a text.