How suvam-09/movie-recommendation-system works
This is a Streamlit content-based movie recommender. A user selects a movie and requests recommendations; the app ranks precomputed similarity scores, resolves the six results against its movie catalog, fetches poster metadata from TMDB, and displays the recommendations. The runtime loads two pickle artifacts. The README describes the TMDB dataset and cosine-similarity approach, but the excerpts do not show how the artifacts are produced, so that offline preparation flow is not drawn.
Main entry points: app.py Similarity ranker · movie_list.pkl Movie catalog · similarity_check.pkl Similarity matrix
Components 8
User experience
- Selection and display
app.py
Recommendation runtime
- Similarity ranker
app.py - Movie catalog
movie_list.pkl - Similarity matrix
similarity_check.pkl
External media
- Poster lookup
app.py
Other components
- Movie seeker
- TMDB movie API
- TMDB image CDN
Connections 8
- Movie seeker → Selection and display: selects movie
README.md - Selection and display → Similarity ranker: requests recommendations
app.py - Similarity ranker → Similarity matrix: reads scores
app.py - Similarity ranker → Movie catalog: looks up titles
app.py - Similarity ranker → Poster lookup: requests posters
app.py - Poster lookup → TMDB movie API: fetches metadata
app.py - Poster lookup → TMDB image CDN: builds image URL
app.py - Selection and display → Movie seeker: displays results
app.py