What this category's data looks like
International student admissions and enrollment management involves materials such as admission brochures, admission letters, and visa application forms. Sources are the admissions system, email receipts, and paper scans, with an update cadence that rolls with application batches and visa progress. Brochures provide fields for major, program duration, fees, and deadlines; admission letters carry name, student ID, admitted major, registration date, and signature; visa application forms record passport number, entry date, signature, and receipt. After materials enter parsing and training tasks, each becomes an independent data row bound to a unique data identifier, for retrieval and subsequent processing to perform name, quantity, and date comparison and discrepancy localization.
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 then completed by retrieval and subsequent processing.
| Object | Check | Verification action |
|---|---|---|
| Admission brochure | Whether fields, dates, and signature are consistent with the admission letter | Check the major name, program duration, fees, and deadline, and confirm the signature or seal is complete; use chunkSize and chunkSplitter to check whether chunk boundaries cut through tables |
| Admission letter | Whether the name, quantity, and date of the same record match across the two materials | Compare name, student ID, admitted major, and registration date, and check the signature and receipt; use trainingType to distinguish training modes |
| Visa application form | Version not synchronized or signature, receipt missing | Check passport number, entry date, and signature, and confirm the receipt is returned; view data row status parsed / indexing / indexed |
| Data row status | In one of the three states and consistent with the interface display | Open the data row to check the status, and confirm that data pending indexing does not participate in recall |
Three common errors
- Symptom: a just-uploaded material cannot be found. Cause: the data row is in parsed or indexing, and data pending indexing is naturally isolated at the retrieval layer, so you must wait for indexing to complete.
- Symptom: the name, quantity, and date of the same record do not match across the two materials. Cause: versions are not synchronized, and the brochure and the letter are each stored independently without cross-material comparison.
- Symptom: empty data rows remain after a training failure. Cause: parsing and chunking have already been stored as visible data rows, and the interrupted training did not write back vectors, so you must edit or delete them directly.
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; the status should be parsed, indexing, indexed, and changing the model rebuilds only data whose indexing is completed; check that the values of chunkSize, chunkSplitter, and trainingType are consistent with the interface.
Business test: use one real record to find a field inconsistent with the ledger, hand it to the reviewer for confirmation, and write it back.
- Check the file record and the data row separately, and confirm that the identifiers of the file and the data row are distinguishable.
- View the data rows covered by the rebuild task, and confirm that only data whose indexing is completed is covered.