Show HN: I built a Movie Recommender using only SQL (opens in new tab)

I wrote a movie recommender using only SQL*

This started out as an experiment to help me find movies to watch based on the vibe of another movie.

For example: you could find movies that are similar to 'Lost In Translation', with less midlife crisis and less Japan but more androids and more future. This should guide to you watch Wong Kar Wai's '2046'.

You don't need to sign-up, there is no tracking, no ads etc.

The approach is directly inspired by Movie Tuner, which is outlined in the Grouplens paper "The Tag Genome: Encoding Community Knowledge to Support Novel Interaction".

*with some Python to present an interface, but the logic is all in SQL