I'm using Lightfm for movie recommendation, my question is quite simple but I can't find the solution to it:
I have trained a model using :
- User Feature matrix (Age, gender etc..)
- Item Feature Matrix (movie category etc ..)
- User Item interactions: if the user watched the movie or not.
So far so good I trained my model, But now I want to predict on a new user who has seen 5 movies.
I have the user features et the new user interaction array but how do I use these information to make a prediction on the other movies he might see.
This is different from a cold start to me because we have the info that the new user has seen 5 movies.
I'm confused because to me it's an expected behavior.
Thanks you in advance for your help!
I heard that we can use partial fit but that bothers me because I want to make a prediction and not train the model on new users
