Event recommender system

A recommender-system project around event affinity, user behavior signals, and retrieval-ready recommendations.

Status: paused · GitHub


What it is

An event-affinity recommender built from user-product interaction histories. The core script pivots long event logs into a user-by-event engagement matrix, inspects event-to-event Pearson correlation, then trains a small embedding model where users and events share a dot-product scoring space.

GitHub