massquantity
commented
9 months ago Hi, there is no direct way of achieving this, but u can use the predict function to get scores after getting the ranked list. The predict function accepts a batch of users/items.
model.predict(user=[1, 1, 1], item=[1, 2, 3])
chiragdaryani
Hi, my training data includes implicit feedback, specifically in the form of past purchases. After training the models, I not only need the ranked list of top n recommendations for each user but also the underlying scores used for ranking.
The function model.recommend_user(user=user_id, n_rec=n) provides the rank but not the scores. Is there a way we can obtain both the scores and the ranked list?