What this category's data looks like
This page addresses vocational college program setup filing materials, including three categories: program setup filing forms, talent demand justification, and faculty and equipment lists. Filing forms are updated by academic year or batch, with fields including program name, code, program length, enrollment target, signature, and date; talent demand justification is updated by regional industry research cycle, with fields including position name, demand quantity, research date, and justification opinion; faculty and equipment lists are updated by semester or equipment procurement milestone, with fields including teacher name, title, equipment name, quantity, date, and receipt. When these three categories of materials enter knowledge base retrieval, the filing form serves as the anchor for program names and codes, talent demand justification serves as supporting evidence for positions and quantities, and faculty and equipment lists serve as condition verification. The retrieval service selects semantic, full-text, or hybrid mode according to material form, and passes candidate chunks and scores to subsequent nodes or humans, to generate handover-ready deployment status and operation records.
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 limit trimming). Handover-ready deployment status and operation records are then completed by subsequent nodes or humans.
| Object | Check | Verification action |
|---|---|---|
| Program setup filing form | Fields, dates, and signatures must be aligned item by item with talent demand justification and faculty and equipment lists; names, quantities, and dates must not contradict each other | In the search test, search using the program name from the filing form, check the number of returned candidates and the similarity score, then switch searchMode to compare candidate differences |
| Talent demand justification | Position names, demand quantities, research dates, and justification opinions must be consistent with the program direction in the filing form, and the version must be the current batch | Search using position keywords, check whether candidates include both semantic and full-text results, record the number of candidates and scores, and verify the justification date and filing batch |
| Faculty and equipment list | Teacher names, titles, equipment names, quantities, dates, and receipts must be consistent with the filing form and justification materials; missing signatures or acknowledgments must be flagged | Search using equipment names, compare the top few rankings before and after enabling usingReRank, check the tokens and array length of the returned reference text, and record separately |
| Deployment status and operation records | Version, connection, and task records must be consistent; the retrieval service connection and task records must be traceable to the same batch | Check item by item once according to the handover form, check which record is missing, and verify whether the reference text after limit trimming retains key original text chunks |
Three common errors
- Symptom: The same program name does not match between the filing form and justification materials. Cause: The two materials were entered in different batches, and fields and dates were not aligned to the same version.
- Symptom: Equipment quantities are inconsistent between the list and the filing form. Cause: After procurement or transfer, only one material was changed, the other was not synchronized to the version, and signatures or acknowledgments are missing.
- Symptom: The same question cannot be recalled when rephrased. Cause:
searchModedoes not match the material form, orusingReRankis not enabled, so candidate ranking is too low.
How to verify the setup
Feature acceptance: Test once each with synonym-rewritten questions, compare the number of candidates and similarity scores, then toggle rerank to observe the top few rankings, and switch searchMode to observe candidate differences.
Business test: Check item by item once according to the handover form, see which record is missing, and verify whether version, connection, and task records are consistent.
- In the knowledge base search test, switch
searchMode, compare the number of candidates and scores across semantic, full-text, and hybrid modes, and record the differences. - Retrieve once before and after enabling
usingReRank, compare the top few rankings, and check the tokens and array length of the reference text afterlimittrimming.