What this category's data looks like
This page addresses the deployment handover scenario for kindergarten operation behavior and quality self-assessment materials, covering three types of materials: self-assessment reports, issue lists, and rectification plans. Self-assessment reports are updated by semester or academic year, with fields including descriptions and conclusions for kindergarten conditions, safety and hygiene, care and education, faculty and staff, and internal management; issue lists are generated per instance, with fields including issue name, quantity, responsible person, and rectification deadline; rectification plans correspond item by item to issues, with fields including measures, completion dates, and signature fields. When these three types of materials enter the knowledge base retrieval and recall function, each is segmented and indexed as an independent source: self-assessment reports provide overall conclusion paragraphs, issue lists provide itemized facts, and rectification plans provide measures and time nodes. During handover, consistency among version, connection, and task records must be used for judgment, and material names, quantities, dates, and signatures must match between the two materials.
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 upper limit trimming). The handover-ready deployment status and operation records are completed by subsequent nodes or manually.
| Object | Check | Verification action |
|---|---|---|
| Self-assessment report | The conclusion paragraphs in the report must be consistent with the names, quantities, and dates in the issue list and rectification plan, and signatures and receipts must be complete | In the search test, use the conclusion sentences from the report to search, check whether the returned candidates cover the corresponding paragraphs, record the number of candidates and scores, and switch searchMode to compare semantic and full-text results |
| Issue list | List entries correspond item by item to the rectification plan; the name, quantity, and deadline of the same issue must not differ in the two places | Use the issue names in the list to search, check whether the returned candidates include both list and plan fragments, and compare the ranking of the top few before and after enabling usingReRank |
| Rectification plan | The measures, completion dates, and signature fields in the plan correspond to the report conclusions, and the version must be synchronized | Use the measure sentences in the plan to search, check the similarity score distribution of the returned candidates, confirm that key fragments are still retained after upper limit trimming, and record the array count separately |
| Deployment operation records | Version, connection, and task records are consistent, and the handover form can be checked item by item | Check the retrieval node connection status and task records item by item according to the handover form, and check whether the number of returned candidates and scores under the limit setting are consistent with the records |
Three common errors
- Symptom: The name or quantity of the same issue does not match between the issue list and the rectification plan. Cause: The two materials were imported separately without association verification, and retrieval candidates hit different sources.
- Symptom: The self-assessment report has been updated, but the issue list and rectification plan are still the old version. Cause: The version was not synchronized, and old fragments are still recalled and ranked at the top during retrieval.
- Symptom: The rectification plan lacks signatures or receipts, and the handover form cannot be checked item by item. Cause: Signatures were not completed during the material circulation process, and the deployment status records are inconsistent with the paper materials.
How to verify the setup
Acceptance of this function: Test once each with synonym-rewritten questions, compare the number of candidates and similarity scores, then toggle usingReRank on and off to observe changes in the ranking of the top few, and switch searchMode to check recall differences.
Business test: Check once item by item according to the handover form, see which record is missing, and focus on checking whether version, connection, and task records are consistent, and whether signatures and receipts are complete.
- In the search test, search once each for the same issue using two different phrasings, record the number of candidates and scores, and confirm that both semantic and full-text results are merged.
- Search once each with rerank enabled and disabled, compare the order of the top few, and confirm that secondary ranking takes effect and that key fragments are still present after upper limit trimming.