Search for the most liked user element with estimated values

Question

Suppose a user votes for some films on a scale of 1 to 5. These films have genre information, and there can be more than one genre in a film. Like this:

Movie A Rating 4 Action/Sci-Fi Movie B Rating 5 Comedy/Action Movie C Rating 4 Comedy/Drama 

We want to know what genre our user likes. Here we have our result set:

 Genre Movie_Count Average_Rating ---------- Action 2 5 Comedy 2 4.5 SciFi 1 4 Drama 1 4 

Obviously, we cannot predict anything with such a small set of results, but suppose we have a larger data set.

Using this data, how can we sort the most popular genres of this user? Just calculate the weighted average or something more complicated?

Answer 1

The main problem that I see here:

Custom bids for 1,000 comedy films with an average score of 4

Custom bids for 10 action movies with an average rating of 4.1

How do you order them?

See http://www.evanmiller.org/how-not-to-sort-by-average-rating.html for a discussion and one possible solution.

Another problem:

If the film is both a comedy and an action, and it was rated 4.0, how much was it because it was a comedy or action?

You can solve this by maximizing expectations http://en.wikipedia.org/wiki/Expectation%E2%80%93maximization_algorithm .