How to get a database for your own recommendation algorithm?

Question

Algorithms

For the sake of interest, I want to write an algorithm for recommending films. Recommendations will be generated based on the genres of those films that the user marks as liked. Also, the cast, director, etc. will play their role. And if you can collect all the information about the films marked by the user, simply by parsing the movie page on some site. Then how to search for films according to the parameters set by the algorithm? Can I get a database of movies somewhere so that the algorithm can run through it and find a suitable movie?

Answer 1

There is a very good API - The Movie Database , it has Russian descriptions and actors and directors, etc. It is free and the API itself has a huge number of examples, besides, initially they can be selected according to the parameters you need.
It is best to save the results of the selection to a local database and simply increase it over time, and subsequently make the main selection from it and supplement it with TMD.

Answer 2

Each criterion is a vote.
Each movie has a stack of votes.
1. For all films, we put down the votes of the selection criterion. (is there an actor needed? - yes: +1; is there such a genre? - yes: +1; etc.)
2. Sort the list in descending order by the number of votes in each stack (ie, in each film).
3. We display a list of films in accordance with preferences.
List of movies on wikipedia

Answer 3

https://www.kaggle.com/c/movie
If you search for "kaggle movie recommendation" or "movie recommendation dataset" you'll get a lot more information.