What this category's data looks like
Vocational school student internship management involves three types of materials: internship agreements, internship plans, and internship logs, which are generated by schools, enterprises, and students at different stages. Agreements contain position names, start and end dates, insurance clauses, and signatures from both parties; plans contain batches, week numbers, rotation arrangements, and instructor confirmation; logs contain daily positions, work hours, mentor signatures, and receipts. When these three types of materials enter the knowledge base retrieval and recall function, agreements and plans are retrieved as institutional basis, while logs are retrieved as process records, each participating in recall as independent fragments. In terms of update cadence, agreements change with batches, plans are adjusted by semester, and logs are submitted weekly. In terms of field characteristics, the names, quantities, and dates of the same record across the three materials must be consistent, version numbers and references must be synchronized, and signatures and receipts must not be missing.
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 limit trimming). The version ledger and change records are completed by subsequent nodes or manually.
| Object | Check | Verification action |
|---|---|---|
| Internship agreement | Whether fields, dates, and signatures are consistent with the current version | In the retrieval node, test once each with synonymous questions, record the number of candidates and scores, and check whether the agreement name, start and end dates, and signature fields are hit simultaneously |
| Internship plan | Whether batches, week numbers, and rotation arrangements match the agreement | Switch searchMode to semantic, full-text, or hybrid, compare whether candidates simultaneously contain plan and agreement fragments, and check whether dates and quantities are consistent |
| Internship log | Whether daily positions, work hours, and mentor signature receipts are complete | Compare the ranking of the top few entries before and after enabling usingReRank, observe whether log fragments are reranked to the front, and check whether signature and receipt fields appear |
| Version ledger and change records | Current version is unique, old versions are traceable, references are synchronized | After changing one piece of content, retrieve again, record changes in the number of candidates and scores, and check whether the referenced text under similarity and limit is updated synchronously |
Three common errors
- Symptom: The position names and dates of the same internship record do not match between the agreement and the plan. Cause: The two materials were imported separately, versions were not synchronized, and each recalled independent fragments during retrieval.
- Symptom: A log missing a mentor signature or receipt is still recalled as a valid record. Cause: Field validation was not completed before ingestion, and the similarity score only reflects text matching.
- Symptom: There are many candidates, but the ones ranked at the front are all irrelevant, and the answer lacks key paragraphs. Cause: The retrieval mode does not match the material form, rerank is not enabled, or similarity filtering is too broad.
How to verify the setup
Acceptance for this function: test once each with synonymously rewritten questions, compare the number of candidates and scores, then toggle rerank on and off to observe the ranking of the top few entries.
Business test: change one piece of content and see whether the version number and references are updated synchronously.
- In the knowledge base search test, switch
searchModeto semantic, full-text, or hybrid, and compare whether the candidate results simultaneously contain semantic and full-text fragments. - Compare the order of the top few entries before and after enabling
usingReRank, and record the tokens and array length of the referenced text undersimilarityandlimit.