What this category's data looks like
Graduate admissions re-examination and admission materials include re-examination plans, candidate score sheets, and adjustment lists. Fields include candidate number, name, preliminary exam total score, re-examination score, total score, major direction, study mode, proposed admission conclusion, date, and signature. Re-examination plans are published by annual batch, score sheets are updated as the re-examination progresses, and adjustment lists are finalized after the adjustment system closes. When materials enter knowledge base retrieval and recall, the re-examination plan provides the rule basis, the score sheet provides per-person scores, and the adjustment list provides proposed admission and vacancy information. Together, the three support authorized access and external services. The update rhythm follows the admissions batch, and fields, dates, and signatures must be checked item by item.
Configuration and implementation checks
This step is executed by the knowledge base retrieval service (retrieval node and search test), outputting candidate chunks and similarity scores (after mode, merge, rerank, filter, and upper-limit trimming). The externally available entry points and authorization scope are completed by subsequent nodes or manually.
| Object | Check | Verification action |
|---|---|---|
| Re-examination plan | Whether the annual batch, major direction, score weights, date, and signature are consistent with the score sheet | In the search test, use searchMode to switch between semantic, full-text, or hybrid, and check whether the plan clauses and score sheet criteria align |
| Candidate score sheet | Whether the candidate number, name, preliminary exam total score, re-examination score, total score, date, and signature match the adjustment list | Test once each with synonymous questions, compare the number of candidate entries and the similarity score, and confirm that the same record's name, quantity, and date are consistent |
| Adjustment list | Whether the proposed admission conclusion, vacancy count, date, and signature are synchronized with the score sheet and re-examination plan | Compare the top few rankings before and after enabling usingReRank, and check that list entries correspond person by person with the score sheet |
| Authorization scope | Whether the visitor's identity matches the visible scope, and unauthorized content is not visible | Access once with a restricted identity, check the number of candidate entries and the cited text, and confirm that unauthorized content is not in the returned results |
Three common errors
- Symptom: The same candidate's name, quantity, or date does not match between the score sheet and the adjustment list. Cause: The two materials were entered by different steps and were not cross-checked by candidate number.
- Symptom: The re-examination plan has been updated, but the score sheet still references old batch clauses. Cause: The version was not synchronized, and historical chunks were mixed into retrieval candidates.
- Symptom: The adjustment list lacks a signature or receipt, but that entry can still be retrieved from the external entry point. Cause: The authorization scope was not narrowed with the material status, making it visible beyond authorization.
How to verify the setup
This feature's acceptance: test once each with synonymously rewritten questions, compare the number of candidate entries and the similarity score, then toggle rerank on and off to observe the top few rankings; record the number of cited entries by array length, and when the citation upper limit is involved, check the tokens of the cited text.
Business test: access once with a restricted identity and see whether the visible scope matches the setting.
- In the search test, switch
searchMode, check whether semantic and full-text candidates appear at the same time, and confirm the multi-route merge result. - Compare the order of the top few entries before and after enabling
usingReRank, and record the number of candidate entries and thesimilarityscore.