sharmaroshan/Drugs-Recommendation-using-Reviews ? reverse-engineered prompt

Build me a notebook project that analyzes the UCI drug review dataset and recommends drugs for a patient based on their condition and review text.

I want it to clean the data, explore the most common conditions and drugs, and show clear visualizations of review patterns, ratings, and sentiment. Then train a model that can use the reviews and condition information to suggest the most suitable drugs and help predict useful things like patient satisfaction or condition category. If there are a few different modeling approaches that work well, try them and compare the results, with one stronger final model if possible.

Please make the notebook easy to follow, with clear steps, comments, and charts so someone can understand the story from the data all the way to the recommendation results. If you need current best practices for text modeling or evaluation, look them up online.