Methods, apparatus, devices, and storage media for aggregating cross-page test questions
By acquiring test question images and using label coding to assess relevance, the problem of low accuracy in automatic clustering of cross-page test question elements was solved, achieving efficient and accurate automatic classification of test question elements.
Patent Information
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- HANGZHOU ZHIJUAN PLANET TECHNOLOGY CO LTD
- Filing Date
- 2026-05-26
- Publication Date
- 2026-06-26
AI Technical Summary
Existing technologies cannot effectively perform automatic clustering of test item elements, especially in the case of cross-page scenarios, where clustering accuracy is low, and manual integration and conventional text matching methods suffer from low efficiency and insufficient accuracy.
By acquiring test question images, extracting text blocks and dividing them into text clusters, evaluating the relevance using label encoding, and combining the trained aggregation model for automatic clustering, the accurate classification of the same test question elements is ensured.
It achieves highly accurate automatic clustering of test question elements, solves the problem of clustering cross-page test question elements, improves clustering efficiency and accuracy, and reduces human error.
Smart Images

Figure CN122290156A_ABST