Leveraging Text Mining for Drug Recommendation System

The advent of social media has significantly empowered patients to share their medication experiences across various online platforms. These reviews reflect diverse sentiments, highlighting the positive and negative effects of the prescribed drugs on their health. Analyzing these user-generated reviews on social media can uncover latent details regarding the efficacy of the drugs, the possible side effects, and the patient's satisfaction level. These reviews also help other stakeholders, like pharmaceutical companies and healthcare professionals, gain valuable insights about the drug. Text mining techniques can be leveraged to examine these reviews and identify their associated sentiments. In this research, we develop a Drug Recommendation System using Machine Learning and Deep Learning models. The sentiments of the patients are analyzed to decide on the most suitable drug for a particular medical condition. The performances of these models were evaluated and compared using metrics - accuracy, precision, recall, and F1 score. Empirical results demonstrate that Bi-directional RNN and Light Gradient Boosting models outperformed other models taken for this study.