Database / Weaviate Vector database Interview questions
What is the alpha parameter in Weaviate hybrid search?
alpha controls the relative weighting between the vector search component and the BM25 keyword component in a hybrid search query, letting an application tune how much each signal contributes to the final fused ranking for a given use case.
collection.query.hybrid(query="running shoes", alpha=0.0) # pure BM25 keyword search collection.query.hybrid(query="running shoes", alpha=0.5) # balanced blend collection.query.hybrid(query="running shoes", alpha=1.0) # pure vector search
An alpha of 0 reduces the query to pure BM25 keyword search, 1 reduces it to pure vector search, and values in between blend the two according to that ratio. There's no universally correct value: content where exact terminology matters heavily (product codes, legal citations) often benefits from a lower alpha favoring keyword matching, while more conversational, paraphrase-heavy content often benefits from a higher alpha favoring semantic vector matching, and many teams tune this empirically against their own query logs and relevance judgments.
More Related questions...
