A legal firm's assistant is missing relevant contract clauses when users phrase questions differently from the exact terminology in the documents. For example, a search for 'termination rights' fails to surface passages about 'right to end the agreement.' Which configuration change addresses this?

Neural semantic search uses dense embedding models to represent queries and passages as vectors in a shared semantic space, so conceptually similar phrases—even when they share no words—are matched by meaning. Keyword search matches on token overlap and cannot bridge terminology gaps like 'termination rights' versus 'right to end the agreement.' Switching to neural semantic search is the appropriate solution. The correct answer, "Switch from keyword search to neural semantic search to enable meaning-based matching", directly satisfies the scenario because it aligns with watsonx Assistant's design principles and the specific capability being tested. The incorrect options ("Manually add synonym lists to the Elasticsearch index mapping", "Increase the number of search results returned (top-K) to retrieve more candidates", "Enable fuzzy character-level matching in the search integration settings") may appear relevant but each misses a key requirement or introduces a step that is either unnecessary or belongs to a different workflow. Mastering the distinction between these approaches is essential for effective watsonx Assistant implementations and is a core focus of the Domain 4: Integrate with watsonx section of the certification exam.