Unverified50% confidenceFactExact time
Recall@k and MRR are commonly used to evaluate retrieval quality, while evaluating generation typically looks at the consistency between answers and sources, i.e., faithfulness.
1
Sources
50%
Confidence
Long-term
Relevance
10/12/2026
First Seen
Sources
Related Entities
Related Claims
UnverifiedKnowledge base evaluation focuses on retrieval accuracy, recall, faithfulness (whether generated results are supported by reference materials), and the collaborative effect of each component.70% similarUnverifiedBuilt-in evaluators cover general quality dimensions, such as the helpfulness, relevance, and correctness of answers, and are ready to use out of the box.69% similarUnverifiedEvaluation measures output quality through predefined test sets and scoring criteria, with dimensions including accuracy, relevance, faithfulness, and harmful content detection.66% similarUnverified可溯源生成领域的评估标准包括归因精度（attribution precision）和归因召回率（attribution recall）66% similarUnverifiedThe quality validation stage requires identifying missing values, outliers, and logical contradictions, and recording data quality scores to inform downstream analytical users of data reliability.66% similar
Cite This Claim
Stable URI
https://kongchang.com/claim/1010480API
curl https://kongchang.com/api/v1/knowledge/claims/1010480MCP
get_claim(id=1010480)