Supervised Machine Learning for Recommendation of Drugs and Sentiment Rating

Abstract

The various diseases attacking the human body, such as the coronavirus and so on; nowadays, due to the increase in infections, there are no systems and medical experts so that patients can take medicines at their own risk. Still, they cause severe damage to the patient's body and cause death. To solve that problem, the author introduces the drug recommendation system based on machine learning and sentiment in this paper. They can take the name of the disease from the patient, recommend the drug for the given condition, and provide the SENTIMENT based on the experience of earlier users. If the rating is high for the predicted disease, then patients recommend and trust the drug. The TF-IDF (Term frequency-inverse document frequency) algorithm is used to extract features. We use different machine learning algorithms to determine accuracies, such as the SGD classifier, Multilayer perceptron classifier, Nave Bayes, Ridge classifier, Linear SVC, Logistic Regression, WORVEC, and BAG of WORDS. These features extracted will be applied to the different machine learning algorithms. We use the TF-IDF feature extraction algorithm among all the algorithms because it performs best. The UCI machine learning website used a DRUG REVIEW dataset to implement the project.