Movie Recommender System
View on GitHubMachine Learning
Movie Recommender System
A content-based movie recommendation project that represents movie metadata as text features and ranks similar titles with cosine similarity.
PythonPandasNumPyScikit-learnCosine Similarity
Key Highlights
- Builds movie representations from genres, cast, keywords and descriptions.
- The earlier claims about collaborative filtering, 100K+ interactions and a 23% improvement are not supported by the current README and have been removed.
Tech Stack
PythonPandasNumPyScikit-learnCosine Similarity
Overview
A content-based recommender that compares movie metadata to find titles similar to a selected movie.
How It Works
- Extract features including genres, cast, keywords and movie descriptions.
- Vectorize the text representation and calculate cosine similarity.
- Return titles with similar metadata to the user's selected movie.
Tech Stack
- Python, Pandas, NumPy and scikit-learn.
Getting Started
The README describes running the application and entering a movie title to receive recommendations. The repository does not document a reproducible setup command.