an area of supervised machine learning in artificial intelligence. It is closely related to regression and classification, but the goal is to learn from examples a similarity function that measures how similar or related two objects are
Discovered by embedding cosine similarity (sentence-transformers MiniLM, 384-dim).