What this category's data looks like
In the academic journal submission initial review scenario, three types of materials—submitted manuscripts, author information forms, and plagiarism check reports—enter the document parsing and chunking stage simultaneously. Submitted manuscripts carry multilingual mixed content of the main text and references, author information forms carry structured fields such as names, affiliations, and authorship order, and plagiarism check reports provide comparison conclusions and dates. The material sources are author submissions and editorial office registration, and the update cadence rolls with initial review rounds, so the same manuscript may have its version replaced multiple times. In terms of field characteristics, manuscripts contain headers and footers, footnotes, and cross-page tables; author information forms contain Chinese-English bilingual names and institutional translated names; and plagiarism check reports contain report numbers, detection dates, and signature fields. When entering this feature, manuscripts and reports are first split into retrievable data rows, author information forms are used for back-reference and cross-checking, and the three together support the output of the unified terminology table and comparison draft, which is submitted for checking before external release.
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 "unified terminology table and comparison draft" is then completed by retrieval and subsequent processing. After parsing is completed, parsing and chunking 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 |
|---|---|---|
| Submitted manuscript | If the name, quantity, or date of the same record does not match across the two materials, verify whether the manuscript version is consistent with the author information form and plagiarism check report | Check the file records and data rows separately, and check whether the status of each data row after splitting by chunkSize and chunkSplitter is parsed, indexing, or indexed |
| Author information form | Author names and institutional translated names must be able to back-reference the original text, and must match the authorship order in the manuscript and the registration information in the plagiarism check report | Open the data rows to check the status, sample three terms to check whether their spelling is consistent throughout the whole text, and check the processing mode corresponding to trainingType |
| Plagiarism check report | The report number, detection date, and signature or receipt must be complete; if the version is not synchronized, it must not enter external release | Look at the data rows covered by the rebuild task, confirm that only data with completed indexing is covered, and that data pending indexing does not participate in the rebuild |
| Unified terminology table and comparison draft | The same term must be consistent throughout the whole text, translated names must be able to back-reference the original text, and there must be no conflict with the fields of the three materials | Sample three terms to check whether their spelling is consistent throughout the whole text, and record the number of citations and the tokens budget of the cited text |
Three common errors
- Symptom: materials just uploaded cannot be found. Cause: the data rows are in the parsed or indexing state, and data pending indexing is naturally isolated at the retrieval layer; successful parsing does not mean retrieval will definitely hit.
- Symptom: the name, quantity, or date of the same record does not match across the two materials. Cause: the manuscript version is not synchronized, and the author information form and plagiarism check report still point to the old version, resulting in two sets of spellings in the terminology table and comparison draft.
- Symptom: empty data rows are left after training fails. Cause: training interruption leaves orphan data, materials lacking signatures or receipts are not intercepted, and the rebuild scope and the initial training coverage scope are mixed up.
How to verify the setup
This feature's acceptance: upload one material, observe the status transitions of data rows and the coverage scope of the rebuild task, and confirm that the status transitions among parsed, indexing, and indexed and is consistent with the interface display.
Business test: sample three terms to check whether their spelling is consistent throughout the whole text, check whether translated names can back-reference the original text, and record the number of citations and the tokens budget of the cited text.
- Open the data rows to check the status, check the file records and data rows separately, and confirm that the identities of the file records and data rows are distinguishable from each other.
- View the data rows covered by the rebuild task, confirm that only data with completed indexing is covered, and check whether
trainingTypecorresponds to the processing mode.