What this category's data looks like
This page addresses primary and secondary school mental health screening and referral scenarios. Materials come from three categories: screening plans, interview records, and referral forms, with multilingual versions and terminology alignment requirements. Screening plans provide scale names, screening dimensions, and judgment criteria; interview records carry student statements, teacher observations, and referral recommendations; referral forms record the receiving party, date, and signed receipt. When these three categories of materials enter this feature, screening plans and interview records serve as terminology sources and contextual references, while referral forms serve as the reference for field and signature verification. Materials are updated according to the semester screening cycle and the pace of individual case referrals, commonly involving multilingual parallelism, version iteration, and supplementary receipt entry. Field characteristics include student identifiers, screening tool names, dimension score ranges, interview dates, referral destinations, receipt confirmations, and signature fields. The same term must remain consistent in wording across different languages and different materials and must be traceable back to the source text.
Configuration and implementation checks
This step is executed by the indexing service and vector database, outputting vectors and index status (whether recall is possible, whether data in the same group is merged). The "unified terminology table and alignment draft" is then completed by subsequent AI nodes, code nodes, or manual work (retrieval and citation trimming only take supporting materials).
| Object | Check | Verification action |
|---|---|---|
| Psychological screening plan | The same scale name and dimension terminology are consistent in wording throughout the multilingual materials, and the translated name can be traced back to the source text; screening dimensions and judgment criteria do not become ambiguous due to language versions | In the knowledge base settings, verify that the indexing model is consistent with the vector model actually used by the knowledge base; open a screening plan data row and confirm that the index field records multiple vector identifiers with a status of indexed |
| Interview record | When the name, quantity, or date of the same record does not match between two materials, the original text of the interview record must be used as the basis for tracing back; old drafts whose versions are not synchronized do not enter the externally published alignment draft | Search once by student identifier and interview date to check whether data in the same group is merged and whether the vector score takes the highest score; verify the citation text budget corresponding to limit, and record the number of array entries separately |
| Referral form | The receiving party, referral date, and signed receipt fields of the referral form are complete; records missing a signature or receipt do not enter the externally published alignment draft | Open the referral form data row and verify the index field; check the data range covered by the rebuild task records, and confirm that after changing the model, rebuilds are triggered only for data that has already been indexed |
Three common errors
- Symptom: duplicate entries appear in retrieval results. Inspect the following: the same group of source data corresponds to multiple vector identifiers, and when merging back to the source data, they are not merged by the highest score, causing duplicate presentation of content from the same group.
- Symptom: the name, quantity, or date of the same record does not match between two materials. Inspect the following: the versions of the screening plan, interview record, and referral form are not synchronized, and old drafts and new drafts are ingested in parallel.
- Symptom: referral forms missing a signature or receipt still enter the alignment draft. Inspect the following: field verification only takes supporting materials and does not set up a separate check for the signature field and receipt status.
How to verify the setup
Feature acceptance: take one data row and look at the index field to confirm that multiple vector identifiers are all present and the status is indexed, then search once to see whether the same group is merged and whether the score takes the highest value.
Business test: take three terms and check whether their wording is consistent throughout the text, and verify whether the translated names can be traced back to the source text.
- In the knowledge base settings, verify that the indexing model is consistent with the vector model actually used by the knowledge base, and check the data range covered by the rebuild task records.
- Open a data row and verify the index field to confirm that multiple vector identifiers are all present; during retrieval, record the number of citations as the array length, and verify the tokens of the citation text separately using
quoteMaxTokenas the citation content budget.