I am looking to implement a small recommendation engine for a VoD movie database. I am looking at the following type of recommendation:
When the customer is already looking at a certain movie, I would like to make a recommendation like Amazon's "Customers Who Bought This Item Also Bought...".
The data i have available to me is all customer's previous purchases and metadata for the movies (genre and such).
Given this information, which algorithms or methods would be the most effective?
My first (perhaps naive) idea is, given the movie the customer is currently looking at, to find all the other customers who have bought this movie and and among these customers to find the movies they have bought the most times.
Got something better?
[link] [13 comments]