Glossary term
Glossary term
Memory and Retrieval
In recommendation systems, a matrix of embedding vectors generated by matrix factorization that holds latent signals about each item. Each row of the item matrix holds the value of a single latent feature for all items. For example, consider a movie recommendation system. Each column in the item matrix represents a single movie. The latent signals might represent genres, or might be harder-to-interpret signals that involve complex interactions among genre, stars, movie age, or other factors.
The item matrix has the same number of columns as the target matrix that is being factorized. For example, given a movie recommendation system that evaluates 10,000 movie titles, the item matrix will have 10,000 columns.
Created for this library
A streaming recommender team factorizes its rating matrix into a user matrix and an item matrix whose rows represent each movie's latent factors.
An e-commerce recommender stores the item matrix of product latent factors as a precomputed table for fast similarity lookup.
A music platform stores the item matrix of song embeddings so similar songs can be retrieved by inner product for related-track widgets.
Definition source: Google for Developers Machine Learning Glossary | Creative Commons Attribution 4.0 License