What this category's data looks like
School moral education and homeroom teacher work materials include themed class meeting plans, student comprehensive quality records, and home visit records. Their sources are scattered across grade groups, homeroom teachers, and the moral education office, and updates are triggered as semesters progress and special activities occur. The fields involve activity name, date, number of participants, signatures, and receipts. When these materials enter knowledge base retrieval, themed class meeting plans serve as the basis for activity design, student comprehensive quality records serve as process evidence, and home visit records serve as evidence of home-school communication; each of the three plays its own role in recalled snippets, and the retrieval node uniformly returns candidates, based on which subsequent nodes or humans form a version ledger and change records.
Configuration and implementation checks
This part is executed by the knowledge base retrieval service (retrieval node and search test), outputting candidate snippets and similarity scores (after mode, merge, rerank, filter, and upper-limit trimming). The version ledger and change records are then completed by subsequent nodes or humans.
| Object | Check | Verification action |
|---|---|---|
| Themed class meeting plan | Check whether the activity name, date, number of participants, and signature column are consistent with the class records. When the name, quantity, or date of the same plan do not match across two materials, the current version prevails. | In the knowledge base search test, switch searchMode to semantic, full-text, or hybrid, and compare candidate snippets and similarity scores; check the ranking of the top few items before and after enabling usingReRank. |
| Student comprehensive quality record | Check whether the record name, quantity, date, and signature or receipt are complete. When the version is not synchronized, the old version can be traced, and the current version is unique. | Test once each with synonym-rewritten questions, and observe the number of candidates and scores; record the length of the citation array, and separately check whether the tokens of the cited text exceed the limit trimming. |
| Home visit record | Check the home visit date, student name, home visit teacher signature, and parent receipt. If the signature or receipt is missing, mark it as pending supplementation. | In the search test, check whether the returned candidates include both semantic and full-text results, and confirm that key snippets of the original text are still retained after multi-path merging. |
| Version ledger and change records | Check that the current version is unique, the old version can be traced, and citations are synchronized. When the name, quantity, or date of the same record do not match across two materials, a change is triggered. | After changing one piece of content, retrieve again to see whether the version number and citations are updated synchronously; compare the number of candidates and score changes before and after similarity filtering. |
Three common errors
- Symptom: The name and date of the same themed class meeting plan do not match across two materials. Cause: The version is not synchronized, and the old version and the new draft coexist without the current version being marked.
- Symptom: The student comprehensive quality record lacks a signature or receipt. Cause: The material circulation process is not closed-loop, and the signature page is not archived together with the record.
- Symptom: The home visit record cannot be recalled when the question is phrased differently. Cause: Question optimization does not cover references, and the retrieval mode does not match the material form.
How to verify the setup
Acceptance of this feature: test once each with synonym-rewritten questions, compare candidate snippets and similarity scores, then toggle rerank on and off to observe the ranking of the top few items; record the number of candidates and scores, record the length of the citation array separately, and check the cited text tokens and limit trimming separately.
Business test: change one piece of content and see whether the version number and citations are updated synchronously; if the number of cited items changes, check the similarity filter and searchMode settings, and confirm the candidate sources and hit associations.
- In the knowledge base search test, switch
searchModeto semantic, full-text, or hybrid, compare candidate snippets and similarity scores for the same question, and record the number of candidates. - Enable or disable
usingReRank, and compare changes in the ranking of the top few candidates; after adjustingsimilarityandlimit, separately record the length of the citation array and the cited text tokens.