A Novel Design of Deep Learning Assisted Drug Recommendation Model using Sentimental Inspection

The field of Natural Language Processing includes Sentiment Analysis as one of its core algorithms for identifying emotional tone in written content. As a result of the corona virus, many people have died, and the medical community as a whole is in shock. As a result of the shortage, many began self-medicating without first consulting with a professional, worsening their health. In recent years, machine learning has proven useful in a wide range of contexts, and novel efforts have been made to automate formerly laborious tasks. Our group decided to focus on studying medicine evaluations written by others, with each review including a star rating from 1 to 10. Since there are several reviews for pharmaceuticals that treat the same medical problem, we thought it would be interesting to see if the reviews for these ailments use the same terms or if the conditions' specific vocabulary choices affect the products' overall scores. To that end, we set out to construct supervised ML categorization techniques that can deduce the rating's category from the review's content. Our primary focus had been on implementing various embedding’s, including Term Frequency Inverse Document Frequency (TFIDF). Accuracy, recall, f1score, reliability, and AUC were used to assess the anticipated emotions.