movie_Recommendation

Movie Recommendation System Overview

This repository implements movie recommendation systems using clustering, content-based filtering, and collaborative filtering techniques. Each system offers unique approaches to recommend movies based on user preferences and movie content.

For using this systems please go to the Hugging Face Space.

Clustering-Based Recommendation System

This system clusters users based on their movie ratings using KMeans and PCA. It identifies optimal clusters using the elbow method, normalizes ratings, and recommends movies based on user preferences within the same cluster.

Content-Based Recommendation System

This system recommends movies based on content similarity. Features like title, overview, genre, and director are processed into a "soup" using TF-IDF. Recommendations are made using cosine similarity and displayed through Streamlit with detailed movie information.

Collaborative Filtering Recommendation System

This system recommends movies by analyzing user ratings sing collaborative filtering. It calculates user similarity with cosine similarity, selects similar users, and suggests movies based on their preferences. The Streamlit interface allows users to rate movies and receive personalized recommendations.