Vector Models and Indexes for Self-Study Exam Records and Graduation Review: Find Supporting Evidence

Self-Study Exams: review vector indexes for evidence lookup. Check source versions, input mappings and the values received by downstream nodes.

What this category's data looks like

This page addresses the self-taught examination registration records and graduation review business. Materials include registration records, course pass certificates, and graduation application forms. Registration records capture the candidate's basic information and enrollment status; course pass certificates capture individual subject scores and the date obtained; graduation application forms capture the graduation application and review signatures. The three types of materials are continuously updated by exam session and batch, with fields covering name, exam registration number, course name, score, date, and signature fields. When entering the vector model and index feature, registration records serve as the basis for identity and enrollment status, course pass certificates serve as the basis for courses and scores, and graduation application forms serve as the basis for graduation conclusions and approvals. Materials often show inconsistencies in the same record's name, quantity, or date, unsynchronized versions, or missing signatures and receipts, requiring version reconciliation and field alignment before indexing.

Configuration and implementation checks

This step is executed by the index service and vector database, outputting vectors and index status (whether recall is possible, whether data in the same group is merged). "Basis entries that can be traced back to the original text" are then completed by subsequent AI nodes, code nodes, or manual work (retrieval and citation trimming only take supporting materials).

ObjectCheckVerification action
Registration recordsName, exam registration number, enrollment status, and date must be aligned item by item with course pass certificates and graduation application forms; signatures and receipts must not be missingOpen a data row to check whether the index field records multiple vector identifiers, and confirm that the row status is indexed
Course pass certificatesCourse name, score, and date obtained must be consistent with registration records and graduation application forms, and the version must be currentCheck that the index model is consistent with the vector model actually used by the knowledge base, and review the data scope covered by the rebuild task records
Graduation application formsGraduation conclusion, approval date, signatures, and receipts must correspond to registration records and course pass certificates; names and quantities must not conflictInitiate a retrieval for the same record and check whether the recall results are merged into the same group of data and the highest score is taken
Data rowMultiple vector identifiers for one piece of data must be complete; after changing models, the rebuild only covers data that has already been indexedCheck the reference budgets related to limit; array counts are recorded separately and are not compared for equality

Three common errors

  • Symptom: duplicate entries appear in retrieval results. Inspect the following: the name or date of the same record does not match across two materials, and the vectors 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 records do not cover all data rows that have already been indexed.
  • Symptom: a data row has content but cannot be retrieved. Inspect the following: the index field is missing vector identifiers, or missing signatures and receipts cause the material not to enter the index.

How to verify the setup

This feature's acceptance: sample a data row to see whether the index field records multiple vector identifiers, then run a retrieval to see whether data in the same group is merged and whether the highest score is taken.

Business test: ask a basis question that spans materials and see whether the answer provides a traceable source.

  • In the knowledge base settings, check that the index model is consistent with the vector model actually used by the knowledge base, and review the data scope covered by the rebuild task records.
  • Open a data row to check the index field, confirm that the status is indexed, and check the reference budgets related to limit; array counts are recorded separately.