Movie Recommendation System

This project is a simple user-based and item-based collaborative filtering movie recommendation system built with Python and NumPy. It predicts user ratings and recommends top-N mock movies to users based on their past preferences and item similarity.

Features

• Predict missing ratings using item-based collaborative filtering
• Recommend top-N movies for a given user
• Identify the best user to recommend a specific movie to

Files

• recommendation.py — Main logic for predictions and recommendations
• data.csv — Sample user-item rating matrix (not included, add your own)
• README.md — Project overview and usage instructions

Key Functions

• predict_item_based(user_idx, item_idx, k=2)
  Predicts a user's rating for a movie using the k most similar items.

• recommend_top_n_movies(user_idx, ratings, n=3, k=2)
  Recommends the top-N movies a user is most likely to enjoy.

• find_best_user_for_movie(item_idx, ratings, k=2)
  Finds the user who would most likely enjoy a specific movie.

How to Use

1. Prepare your ratings matrix as a NumPy array with shape (num_users, num_movies).
2. Call the relevant functions with user/movie indices.
3. Adjust k (number of neighbors) to tune prediction quality.

Example

user_idx = 0
recommend_top_n_movies(user_idx, ratings, n=3)