Welcome to ImplicitMF’s documentation!

ImplicitMF is a Python package that generates personalized recommendations for implicit feedback datasets. Unlike explicit feedback (e.g., movie ratings), implicit feedback looks at a user’s interactions with an item and uses this as a surrogate measure of their preference toward that item.

ImplicitMF provides a selection of implicit-specific matrix factorization techniques to generate item recommendations for a set of users:

• Alternating Least Squares: as described in Collaborative Filtering for Implicit Feedback Datasets See implicit package for more information on its Python
• Learning to Rank: as described in BPR: Bayesian Personalized Ranking from Implicit Feedback and WSABIE: Scaling Up To Large Vocabulary Image Annotation. See LightFM package for more information on its Python implementation.

Developer Documentation:

• Transform
• Pre-process
• Validation
• Post-process