Give it a movie you love, and it suggests the ones you'll probably love next.
A recommendation engine that suggests movies based on the ones you already enjoy. It's built on the MovieLens dataset — over 100,000 ratings across roughly 9,700 films from 610 users — and uses similarity-based collaborative filtering to find titles that closely match your taste.
Given a movie you like, it measures how similarly other viewers rated it against every other film, then ranks the closest matches — surfacing recommendations that people with overlapping taste tended to rate highly.
100k+ ratings across ~9,700 movies as the foundation.
Ranks films by how alike their rating patterns are.
Learns taste from what like-minded viewers enjoyed.
The Python stack that powers the modelling.
binod