What this category's data looks like
In the scenario of signing scientific research project task books, the materials mainly consist of the project task book, the subject task breakdown table, and the budget book; the fields, dates, and signatures of the three must be consistent. The project task book provides the overall objectives, period, and responsible entities; the subject task breakdown table breaks tasks down to subjects and milestones; the budget book carries the funding categories and amounts. The materials come from the final versions submitted by the scientific research management department and the research group, and the update cadence follows the signing and change process; signatures and receipts must be complete. When entering the knowledge base retrieval and recall function, the project task book and budget book provide the basis for the authorization scope and visible fields required for external services, while the subject task breakdown table supplements the task hierarchy and milestone dates; together, the three determine the source of candidate snippets for matching visitor identity and visible scope. Common inconsistencies are concentrated in the same record having mismatched names, quantities, and dates across two materials, versions not being synchronized, and missing signatures or receipts.
Configuration and implementation checks
This step is executed by the knowledge base retrieval service (retrieval node and search test), outputting candidate snippets and similarity scores (after pattern, merge, re-rank, filter, and upper-limit trimming). The externally available entry points and authorization scope are then completed by subsequent nodes or manually.
| Object | Check | Verification action |
|---|---|---|
| Project task book | Fields, dates, and signatures must be aligned item by item with the subject task breakdown table and budget book | In the knowledge base search test, switch searchMode and compare the candidate snippets returned by the semantic, full-text, and hybrid modes, checking whether names, quantities, and dates are consistent with the budget book |
| Subject task breakdown table | Task hierarchy, milestone dates, and responsible entities must be consistent with the project task book, and versions must be synchronized | Compare the order of the first few candidates before and after enabling usingReRank, and check whether the breakdown table milestone dates match the project task book period |
| Budget book | Funding categories, amounts, and signature receipts must be consistent with the project task book and breakdown table | Adjust similarity and limit, observe changes in the number of candidates and scores, and check whether budget category names and quantities are consistent with the task book |
| External service entry point | Visitor identity and visible scope must match, and unauthorized access must not be visible | Access once with a restricted identity, see whether the visible scope matches the setting, and record the cited text tokens and array item count separately |
Three common errors
- Symptom: The same subject name is written differently in the project task book and the breakdown table, and retrieval candidates do not match. Cause: The two materials have not been field-aligned, and names and quantities are not unified.
- Symptom: The budget book version is not synchronized, and the date and signature fields are missing. Cause: After a change, only one material was updated, the receipt was not collected, and the version was not synchronized.
- Symptom: There are many candidates, but the ones ranked first are all irrelevant, and the answer lacks key passages. Cause: The retrieval mode does not match the material form, re-ranking is not enabled, or similarity filtering is too broad.
How to verify the setup
Acceptance of this function: Test once each with synonym-rewritten questions, compare candidates and scores, then toggle re-ranking on and off to see the ordering of the first few items; switch searchMode and adjust similarity and limit, recording the number of candidates and scores.
Business test: Access once with a restricted identity, and see whether the visible scope matches the setting.
- In the knowledge base search test, retrieve the same question using semantic, full-text, and hybrid modes respectively, compare candidate snippets and similarity scores, and check whether names, quantities, and dates are consistent with the budget book.
- Retrieve once each with re-ranking enabled and disabled, compare the order of the first few candidates; adjust similarity filtering and the citation upper limit, check the cited text tokens, and record the array item count separately.