What to expect: This website lets you rate movies and accordingly generates recommendations for you.
How it works:
1. Login/Register your account through "Sign In"
2. Find a movie from the "Movies" section or use the "Search" feature
3. Add review (rating &/ comment) on a movie page
4. Repeat steps 2 and 3, if needed.
5. Hit "Movie Suggestions" on the bar and view your recommendations (Refreshing the page displays new items, but from the same cluster).
How the Recommendations feature works: It's a Machine Learning algorithm which groups (clusters) you with users who have rated similarly as you.
Other Features:
- CRUD operations (Add/Edit/Delete Review, with user authentication)
- User Management (Login, Registration)
- KMeans Clustering algorithm (Recommendations feature with Collaborative Filtering)
- Search Suggestions and Sort By options (Movies)
- Responsive Layout (Device-compatible web pages)
- Batch process for updating clusters (/reviews/management/commands/runUpdClusters.py)
Technologies:
- Python 2.7 with Django Framework
- Web Templates: HTML, JavaScript, CSS, Bootstrap
- Database: PostgreSQL (Migrated from SQLite)
- Deployment: Heroku
- Data Load: Pandas
Dataset (CSV): MovieLens (ml-latest-small)
- 100,000 ratings and 1,300 tag applications applied to 9,000 movies by 700 users. Last updated 10/2016.
- Files (/data): Users, Movies, Reviews, Genres
References: