What this category's data looks like
This category's data looks like this In the student scholarships, grants, loans, and financial hardship determination scenario, three types of materials are involved: the family financial hardship determination form, the scholarship/grant application form, and the loan receipt. The sources are the school financial aid center, departments, and the handling bank, and they enter in batches along with the academic-year determination, batch review, and disbursement nodes. The three types of materials have different roles: the determination form carries the family economic status and the democratic evaluation conclusion, the application form carries the scholarship/grant program and amount tier, and the loan receipt carries the contract number and arrival confirmation. Field characteristics are concentrated in name, student ID, amount, date, signature, and receipt number, and the same record needs to be aligned between two materials. Material versions are updated with the determination year and review round. When old and new versions coexist, it is easy for names, quantities, and dates to fail to match, or for signatures or receipts to be missing. Document parsing and chunking must be completed first, then verification and discrepancy localization can proceed.
Configuration and implementation checks
This step is executed by parsing and training tasks (server-side tasks), outputting data rows and index status (parsed, indexing, indexed). The "discrepancy list" is completed next by retrieval and subsequent processing. After parsing is completed, parsing chunks are immediately persisted as visible data rows and bound to a unique data identifier. The index transitions through the three states parsed, indexing, and indexed, and data pending indexing does not participate in recall.
| Object | Check | Verification action |
|---|---|---|
| Family financial hardship determination form | Name, student ID, number of family members, determination date, and signature must be consistent with the application form | Verify the fields and signature, confirm the data row status is parsed or indexed, and check whether chunkSize and chunkSplitter affect field splitting |
| Scholarship/grant application form | Name, amount tier, review date, and signature must be consistent with the determination form and receipt | Compare the amount and date, confirm the index status corresponding to trainingType, and check whether data pending indexing is isolated |
| Loan receipt | Contract number, arrival amount, disbursement date, and receipt seal must be consistent with the application form | Verify the receipt number and arrival date, confirm the data row is bound to a unique data identifier, and that the rebuild scope covers only indexed data |
| Data rows and index status | Is in one of the three states parsed, indexing, indexed and consistent with the interface display | Open the data row to verify the status, verify the file record and the data row separately, and confirm that changing the model rebuilds only indexed data |
Three common errors
- Symptom: the name, amount, and date of the same record do not match between the determination form and the application form. Cause: the two materials belong to different entry batches, the fields are not aligned to the same object, and the discrepancy is not localized during the verification step.
- Symptom: the old version of the determination form and the new version of the application form exist at the same time, and the versions are not synchronized. Cause: materials are updated with the academic year but are not version-tagged, so after parsing the data rows are mixed together, and old-version content is recalled during retrieval.
- Symptom: the loan receipt lacks a signature or the receipt number is missing, yet it still enters the verification process. Cause: the scanned material is incomplete, the seal field is not validated after parsing and chunking, and the discrepancy list does not mark the missing item.
How to verify the setup
Acceptance for this feature: upload one material and observe the data row status transitions and the coverage scope of the rebuild task. Data rows should go through the three states parsed, indexing, and indexed, data pending indexing should not be recalled, changing the model should rebuild only indexed data, and verify the processing mode and progress corresponding to chunkSize, chunkSplitter, and trainingType.
Business test: use one real record to find fields inconsistent with the ledger. After parsing the determination form, application form, and receipt separately, align them by name and student ID, locate inconsistencies in name, quantity, date, signature, or receipt number, form a discrepancy list, hand it to the reviewer for confirmation, and write it back.
- Open the data row to verify the status, confirm it is in one of parsed, indexing, indexed and consistent with the interface display, and verify the file record and the data row identifier separately.
- View the data rows covered by the rebuild task, confirm that only indexed data is covered, that data pending indexing does not participate in the rebuild, and verify the tokens of the referenced text and the number of array entries separately.