What this category's data looks like
This page focuses on journal three-review three-proof and editorial quality materials, sourced from the editorial office's archived three-review forms, proofs, proofreading records, and author receipts, updated in sync with each issue's finalization and version replacement. The three-review form carries the fields, dates, and signatures of the initial review, re-review, and final review; the proof carries the proof round, page numbers, revision marks, and receipts; the proofreading records carry error types, quantities, and responsible persons. When the materials enter the vector model and index feature, the three-review forms and proofreading records are written as source data into the business database data table, with the index field recording multiple vector identifiers; image descriptions or image vector indexes in the proofs are also generated when model capabilities support it. The index model must be consistent with the vector model actually used by the knowledge base; after changing models, rebuild is triggered only for data that has already been indexed. Field characteristics include mismatches in names, quantities, and dates for the same record across the two materials, versions not synchronized, and missing signatures or receipts, all of which affect whether basis entries can point back to the original text.
Configuration and implementation checks
This part is executed by the index service and the vector database, outputting vectors and index status (whether recall is possible, whether data in the same group is merged). Basis entries that can point back to the original text are then completed by subsequent AI nodes, code nodes, or manual work (retrieval and citation trimming take only supporting materials). Vectors are stored in the vector database, source data is stored in the business database's data table, and the index field records the corresponding multiple vector identifiers; retrieval first recalls vectors, then merges back to the source data by identifier, taking the highest score for the same group of data.
| Object | Check | Verification action |
|---|---|---|
| Index model | Consistent with the vector model actually used by the knowledge base; after changing models, rebuild is triggered only for data that has already been indexed | Check the index model in the knowledge base settings, view the data scope covered by the rebuild task, and confirm that the defaultToken and maxToken values are consistent with the knowledge base settings |
| Index field of the data row | All multiple vector identifiers of one data item are present, and the status value is indexed or a rebuild task record | Open a data row and check the index field, confirm that all multiple vector identifiers are present, and check whether data in the same group is merged |
| Three-review form | The fields, dates, and signatures of the initial review, re-review, and final review are complete, and names and quantities match the proofreading records | Check the names, quantities, and dates of the three-review forms against the proofreading records one by one, check signatures and receipts, and confirm that defaultToken and maxToken have not truncated key fields |
| Proof | The proof round, page numbers, revision marks, and receipts are complete, and the version is the current version | Compare the dates and page numbers of the proofs and the three-review forms, and confirm that when versions are not synchronized, the current version prevails |
Three common errors
- Symptom: duplicate entries appear in retrieval results. Inspect the following: names, quantities, and dates for the same record across the two materials do not match, and vector identifiers are not merged by the same group of data.
- Symptom: recall quality drops after changing models. Inspect the following: versions are not synchronized, and the rebuild task covers only part of the data that has already been indexed, not all current materials.
- Symptom: a data row has content but cannot be retrieved. Inspect the following: signatures or receipts are missing, the material has not completed indexing, and the status value has not reached indexed.
How to verify the setup
This feature's acceptance: sample a data row and look at the index field, confirm that all multiple vector identifiers are present and the status value is indexed or a rebuild task record; then retrieve once to see whether the same group is merged, taking the highest score for the same group of data. The number of citations is the array length, and quoteMaxToken is the citation content budget; the two are not compared for equality; check the tokens of the cited text, and record the array count separately.
Business test: ask a cross-material basis question, for example whether the final review date of a manuscript is consistent with the number of errors in the proofreading records, and see whether the answer gives a source that can be pointed back to.
- Check the index model in the knowledge base settings, confirm it is consistent with the vector model actually used by the knowledge base, and view the data scope covered by the rebuild task.
- Open a data row and check the index field, confirm that all multiple vector identifiers are present, the status value is indexed or a rebuild task record, and check whether data in the same group is merged.