What this category's data looks like
In the preliminary review stage of academic journal submissions, the common materials are the submitted manuscript, the author information form, and the plagiarism check report. The three come from different sources and update at different rhythms: the manuscript is replaced multiple times with revisions, the author information form changes with authorship and affiliation adjustments, and the plagiarism check report is reissued with each detection batch. Field characteristics center on manuscript number, title, author names and order, affiliation, dates, and signed receipts. When the materials enter knowledge base retrieval, their roles differ: the manuscript serves as the main body of text for semantic and full-text recall, the author information form serves as structured assistance for field comparison, and the plagiarism check report serves as an attachment for similarity fragment recall. Together the three support the generation of the version ledger and change records, making the current version unique, old versions traceable, and citations synchronized.
Configuration and implementation checks
This step is executed by the knowledge base retrieval service (retrieval node and search test), outputting candidate fragments and similarity scores (after mode, merge, rerank, filter, and limit trimming). The version ledger and change records are then completed by subsequent nodes or manually.
| Object | Check | Verification action |
|---|---|---|
| Submitted manuscript | Manuscript number, title, receipt date, revision date, and signed receipt are consistent across the two materials | In the search test, retrieve once each with synonymous questions, record the number of candidates and scores, and check the number and dates in the hit fragments |
| Author information form | Author names, order, and affiliation correspond item by item with the authorship in the manuscript body; missing signatures or receipts are marked separately | Switch searchMode to semantic, full-text, and hybrid once each, and compare whether the author field candidates fall in the same record |
| Plagiarism check report | Report number, detection date, and manuscript version number correspond; old reports are traceable and marked with replacement relationships | Retrieve once each with usingReRank enabled and disabled, observe changes in the ranking of the top few results, and check the report fragments against the manuscript version |
| Version ledger | The current version is unique, old versions retain change records, and citations are updated in sync with the version | After changing one title, retrieve again, see whether the candidate fragments and scores point to the new version, and check the number of results after similarity filtering |
Three common errors
- Symptom: The title or receipt date of the same manuscript does not match across the two materials. Cause: After the manuscript was replaced, the author information form was not updated in sync, and the version was not synchronized, causing record misalignment.
- Symptom: The author information form lacks a signature or receipt, and the plagiarism check report date is earlier than the manuscript revision date. Cause: The material collection step did not backfill, and the old report was not marked with a replacement relationship.
- Symptom: Synonymous questions fail to recall the target fragment, and although there are many candidates, the ones ranked at the top are irrelevant. Cause: The retrieval mode does not match the material form, rerank is not enabled, or similarity filtering is too broad.
How to verify the setup
Acceptance of this feature: test once each with synonymously rewritten questions, compare the number of candidates and scores, then toggle rerank on and off to observe the ranking of the top few results; at the same time switch searchMode and adjust limit, record the array length, and separately check the tokens of the cited text.
Business test: change one piece of content and see whether the version number and citations are updated in sync.
- In the search test, synonymously rewrite the same question, record the number of candidates and scores, and check whether the hit fragments cover the manuscript number, author, and date fields.
- After changing one title or date, retrieve again, check that the current version is unique and old versions are traceable, and check that the number of citations and the tokens of the cited text are recorded separately.