What this category's data looks like
This page addresses the scenario of journal publication licenses and annual verification. The materials include three categories: journal publication licenses, annual verification forms, and rectification reports. The license states the journal name, journal number, sponsoring unit, publishing unit, and validity period; the annual verification form is filled in annually with the journal name, journal number, publishing unit, verification conclusion, and signature field; the rectification report records the issue items, rectification measures, completion date, and receipt. The three types of materials come from different sources and have different update rhythms: the license is reissued after changes, the verification form is submitted annually, and the rectification report is generated as matters arise. When entering the knowledge base retrieval service, the license and verification form serve as comparison benchmarks, and the rectification report serves as the source of discrepancies. Each is recalled in fragment form for subsequent nodes or manual verification of names, quantities, dates, and signatures.
Configuration and implementation checks
This step is executed by the knowledge base retrieval service (retrieval node and search test), outputting candidate fragments and similarity scores (after pattern, merge, rerank, filter, and limit trimming). The discrepancy list is then completed by subsequent nodes or manually.
| Object | Check | Verification action |
|---|---|---|
| Journal publication license | Whether the fields, dates, and signatures are consistent with the verification form | Switch searchMode in the search test, compare candidate fragments and scores, and locate discrepancies in journal name, journal number, and validity period |
| Annual verification form | Whether the name, quantity, and date of the same record match the license | Compare the ranking of the top few results before and after enabling usingReRank, and check whether the verification conclusion and signature field have missing items |
| Rectification report | Whether the rectification measures, completion date, and receipt are consistent with the ledger | Use similarity to filter low-scoring candidates, record the candidate array length by number of citations, and verify dates and receipts one by one |
| Knowledge base retrieval service | Whether the candidate fragments and similarity scores cover the three types of materials | Record the number of candidates and scores, and confirm that key fragments of the original text are still retained after limit trimming |
Three common errors
- Phenomenon: The journal name and date of the same record do not match between the license and the verification form. Cause: The versions of the two materials are not synchronized, and old-version fragments are still recalled.
- Phenomenon: The rectification report lacks a signature or receipt, but the verification form shows it has been closed. Cause: The receipt was not written back, and retrieval only recalls fragments of the report body.
- Phenomenon: There are many candidates, but the ones ranked at the front are all irrelevant. Cause: The retrieval mode does not match the material form, rerank is not enabled, or similarity filtering is too loose.
How to verify the setup
Acceptance of this function: Test once each with synonym-rewritten questions, compare candidate fragments and similarity scores, then toggle rerank on and off to observe changes in the ranking of the top few results, and record the number of candidates and scores.
Business test: Use one real record to find fields inconsistent with the ledger, hand them to the reviewer for confirmation, and write back the discrepancy list.
- In the knowledge base search test, switch retrieval modes, compare candidate fragments and scores, and verify names, quantities, dates, and signatures field by field.
- Compare the ranking of the top few results before and after enabling rerank, record the candidate array length by number of citations, and separately verify the tokens of the cited text.