Hello, I have one question: in the movielens dataset, movies in the same timestamp are considered 1 bundle. So if split like that, each bundle ID is highly distinct leading to the number of users interacting with this bundle ID is very small, only about 1-2 users. In this case, how should we handle it?
Hello, I have one question: in the movielens dataset, movies in the same timestamp are considered 1 bundle. So if split like that, each bundle ID is highly distinct leading to the number of users interacting with this bundle ID is very small, only about 1-2 users. In this case, how should we handle it?