product embedding

A product embedding is a machine learning procedure where products are assigned positions in a space. A product vector represents each product’s position in that space. Similar products are close to each other, while products that are different are far from each other (see Product Embeddings and Vectors).