What this category's data looks like
The materials for the establishment and annual reports of Chinese-Foreign cooperative education institutions usually include the establishment application, cooperation agreement, articles of association, annual report filing records, approval receipts, and signature pages. The establishment application focuses on the institution name, education level, program offerings, and enrollment scale; the cooperation agreement focuses on the names of the parties, contribution amounts, validity period, and signing date; the articles of association focus on the governance structure, decision-making procedures, and amendment records. Annual reports are updated annually and are often archived in parallel with the establishment materials. After the materials enter the parsing and training task, they are first split into retrievable data rows, each row is bound to a unique data identifier, and then flows through the three index states for subsequent retrieval and answer citation.
Configuration and implementation checks
This step is executed by the parsing and training task (server-side task), which outputs data rows and index states (parsed, indexing, indexed). The "basis entries that can be traced back to the original text" are then completed by retrieval and subsequent processing.
| Object | Check | Verification action |
|---|---|---|
| Establishment application | The institution name, education level, program, and enrollment scale fields should correspond item by item with the cooperation agreement and articles of association, and the signing date and signature pages should be complete | Check the names, quantities, and dates in the application and the cooperation agreement separately, record inconsistencies, and trace them back to the original page numbers |
| Cooperation agreement | The names of the parties, contribution amounts, validity period, and signing date must be consistent with the articles of association and annual report records; missing signatures or receipts are regarded as an ineffective version | Check the signature pages and receipts of the agreement, and compare the names, quantities, and dates of the same record in the two materials |
| Articles of association | The governance structure, decision-making procedures, and amendment records should be synchronized with the current version, and old versions must not be used as basis entries | Check the version number and amendment date of the articles of association, confirm the current version, and exclude unsynchronized versions |
| Data rows and file records | The identities of file records and data rows are distinguished separately, and the status is consistent with the interface display | Open the data rows to check the status, verify the file records and data rows separately, and confirm the correspondence of chunkSize, chunkSplitter, and trainingType |
Three common errors
- Phenomenon: The same institution name is written differently in the application and the cooperation agreement, and the quantities and dates also do not match. Cause: The two materials were drafted separately, without field-level comparison and unification.
- Phenomenon: After the articles of association are amended, the old version is still retrieved and the answer cites outdated clauses. Cause: The versions were not synchronized, and the current version and historical versions were not marked distinctly.
- Phenomenon: The cooperation agreement lacks signatures or receipts but is still treated as a valid basis entry. Cause: The completeness of the signature pages and receipts was not verified when the materials were ingested.
How to verify the setup
Acceptance for this feature: upload one material, observe whether the data row status flows through parsed, indexing, and indexed, and check that the rebuild task only covers data rows whose indexing has been completed.
Business test: ask a basis question that spans materials, and see whether the answer provides a traceable source.
- Open the data rows to check whether the status is consistent with the interface display, and confirm that the identities of file records and data rows are separate.
- Check the data rows covered by the rebuild task, and confirm that it targets only data whose indexing has been completed, while data pending indexing does not participate in recall.