1
$\begingroup$

Pretty new to ML so sorry if this question has been answered before.

Dataset of:

  • users (100,000 unique)
  • movies (7000 unique) w/ genre data (action, comedy) and plot summary
  • for each user, a list of their movie ratings (20 million records total)

Goal is to create a model that, upon providing a user's list of movie ratings, can recommend movies he/she may enjoy.

I was thinking I could segment my existing user data into 'groups' based off viewing patterns, then place my new user into one of these, and make recommendations from what the group likes. Can possibly also look at movies w/ similar genre tags that the user enjoys.

What sort of model would be best suited for this? Or would you recommend approaching this problem from a different perspective/methodology?

My limited ML knowledge tells me this requires an unsupervised process, something like perhaps a k-clustering algorithm?

Thanks in advance.

$\endgroup$
  • $\begingroup$ After some research, I believe the first place to start is to look into KNN model, an unsupervised model used frequently in collaborative filtering (the community approach I mentioned above) $\endgroup$ – xii Apr 25 at 20:35
0
$\begingroup$

After some research, I believe the first place to start is to look into collaborative filtering techniques (the community approach I mentioned above). As for movie-specific tags, this is called content-based recommendation.

| cite | improve this answer | |
$\endgroup$

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.