Candidate generation produces a manageable set of possible answers before a more expensive scoring or verification stage. In a recommender system, retrieving a few hundred items from millions lets a detailed ranking model examine each candidate. Candidates may come from several retrieval rules; their union can combine recent, popular and personally relevant items.[ref]
One mechanism represents a query by a vector and each item by a vector , then retrieves large inner products using nearest neighbor search. If the vectors have unit Euclidean norm, maximizing also minimizes . Without normalization those rankings need not agree. A candidate omitted here cannot be restored by a later stage that only reorders the retrieved set, so retrieval coverage and computational cost must be assessed together.
New to topics? Read the docs here!