Source Code Github Mark

Group Movie Recommendations

  1. Input the group's Letterboxd account names or IMDb profile links
    ( e.g. imdb.com/user/ur102028277/ — ratings must be public )
  2. Grab some popcorn while recommendations generate
  3. Decide on which movie to watch or have one randomized
  4. Enjoy!

We have     letterboxd/IMDb account(s)!


About this project

I came across Sam Learner's Letterboxd Movie Recommender a while back. It was awesome! I wanted to build something similar but for groups.

The model finds other Letterboxd reviewers who have similar movie tastes to the group members and extrapolates based on their ratings to predict other movies everyone would like. I found out halfway through that my model had a very similar idea to that of Criticker's Taste Compability Index though I decided not to use a percentile system as it wouldn't account for the inherent quality of the movies watched. Check out Criticker if you are still looking for other recommendations, it's really cool!

Built with: Flask, BeautifulSoup (scraper), Pandas (data analysis), Celery and Redis (delivering progress updates)
Deployed with Gunicorn and Nginx

About me

Hi! I'm James and I love to watch movies. I'm currently on a gap year and will matriculate in 2023. If you have any suggestions or find any errors within the site, do let me know on Reddit!

I'm also looking for an internship opportunity (remote, unpaid). If you know of any relevant Software Engineering / Web Dev / Data Science internship positions, do reach out!