tarini-py/Zee-Movies-Recommendation-System ? reverse-engineered prompt

Build me a movie recommendation notebook for a Zee style streaming app using the MovieLens 1M data. I want it to load the user, movie, and ratings files, do a bit of exploratory analysis, and then compare three ways of recommending movies, one based on similar ratings, one using nearest neighbors, and one using matrix factorization with user and movie details so it can also work for new users or new movies.

Please make it easy to run in a notebook and include simple interactive controls so I can pick a movie or user and see recommendations side by side. I also want a clear train and test setup that respects time order, plus a small evaluation section with things like RMSE and ranking quality. If you need to, look up current docs online for any libraries that help with the modeling or widgets. Keep the outputs readable, with charts that explain the data and the results, and make sure the recommendations feel sensible rather than random.