I am building a user based recommendation engine with Python and LightFM and I am trying to figure out how the score of a user to product is generated from the other users.
Specifically, I am trying to answer the following questions:
1) Which users influence the predicted score of an another user over a product?
2) How can I find these users?