Access to licensed healthcare resources has been more difficult to obtain after the coronavirus was identified. This has the effect of dramatically decreasing availability. This includes not only a dearth of healthcare workers but also of necessary tools and medicines. Many people haverecently passed away, and this is due in large part to the problem that the medical community is facing right now. Due to the drug's limited availability, people began treating their symptoms on their own without consulting a medical professional, worsening their already precarious health conditions. This resulted in the drug's eventual release. As more and more uses for machine learning are discovered, more and more work is being done to automate formerly manual processes. Both tendencies are quite new. The study's overarching goal is to showcase a drug recommender system with the potential to drastically cut down on experts' workloads. In this study, we utilize patient feedback and ratings to create a system for recommending therapeutic interventions. To this end, we employ many different vectorization techniques, such as Bow, TF IDF, Word2Vec, and even manual feature analysis. This system can assist in selecting an appropriate drug for the treatment of an illness by applying a wide range of different classification algorithms. Several metrics, including precision, recall, accuracy, f1score, and area under the curve, were used to assess the predictability of the felt emotions. The findings indicate that the TF-IDF Vectorization-based classifier Linear SVC outperforms the other models significantly with a 93% accuracy rate.