Vocabulary Mismatch

Vocabulary Mismatch is a problem where users and authors use different words for the same concept (e.g., a user searches 'car' but the author writes 'automobile'). In lexical retrieval (BM25) this causes matching failures despite identical intent. Solutions include semantic retrieval (embeddings) and lexical expansion tools that generate synonyms, hyponyms, and meronyms, increasing touchpoints with user queries.

Neural Matching (2018) was Google's first system specifically designed to solve this problem.

In practice, use at least 3-5 synonyms for your Central Entity in every article: alongside 'pool' add 'swimming facility', 'aquatic center', 'swimming venue'. Lexical expansion tools generate these variants automatically. Vocabulary Mismatch explains why keyword SEO alone isn't enough — you must cover the semantic field, not just exact phrases.