Unverified50% confidenceFactExact time
Traditional retrieval evaluation datasets often have only single-digit relevant document annotations per query.
1
Sources
50%
Confidence
Long-term
Relevance
9/15/2026
First Seen
Sources
Related Entities
Related Claims
Unverified语义搜索和关键词搜索单独使用时约每4个结果漏检1个，两者结合漏检率降到约十分之一66% similarUnverifiedA unique index should be added to the OpenID field, which is naturally unique in the database; otherwise, duplicate user records may be created and query efficiency may be reduced.65% similarUnverifiedGlobal queries can answer inductive questions requiring a macro perspective (such as the main themes across an entire document set), a scenario that pure vector retrieval is almost incapable of handling.65% similarUnverifiedIn traditional retrieval evaluation, documents not labeled as relevant are typically treated as irrelevant by default, which systematically underestimates the true performance of retrieval models.64% similar
Cite This Claim
Stable URI
https://kongchang.com/claim/922032API
curl https://kongchang.com/api/v1/knowledge/claims/922032MCP
get_claim(id=922032)