A knowledge information retrieval method, device, equipment, medium and program product
By constructing a correspondence between knowledge information and preset questions, and using a large language model to quickly locate the information supporting the answer, the problem of low efficiency in knowledge information retrieval in existing technologies is solved, and efficient and comprehensive knowledge information retrieval is achieved.
CN122132514APending Publication Date: 2026-06-02INDUSTRIAL AND COMMERCIAL BANK OF CHINA
View PDF 0 Cites 0 Cited by
Patent Information
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- INDUSTRIAL AND COMMERCIAL BANK OF CHINA
- Filing Date
- 2025-08-14
- Publication Date
- 2026-06-02
AI Technical Summary
Technical Problem
Existing knowledge retrieval methods are inefficient, especially in full-text searches, where it is difficult to efficiently locate answers to user questions.
Method used
By constructing a correspondence between knowledge information and preset questions, a large language model is used to determine the questions that the knowledge information can answer. Based on question similarity matching, the information on which the answer is based is quickly located, reducing full-text retrieval and improving retrieval efficiency.
Benefits of technology
It enables efficient retrieval of knowledge information, reduces the time required for full-text searches, improves the speed and comprehensiveness of retrieval, and enhances the accuracy of answers.
✦ Generated by Eureka AI based on patent content.
Smart Images
Figure CN122132514A_ABST
Abstract
This application provides a knowledge information retrieval method, apparatus, device, medium, and program product, which can be applied to the field of artificial intelligence technology, involving the application of large models in knowledge retrieval scenarios, and relating to the field of information security. The method includes: determining a target problem; determining N sets of correspondences between knowledge information and preset problems; wherein any knowledge information contains information for answering the corresponding preset problem; N is a positive integer; the preset problems are determined based on a set of candidate problems associated with the corresponding knowledge information, the set of candidate problems including: candidate problems determined based on similar knowledge information of the associated knowledge information, the information similarity between the similar knowledge information and the associated knowledge information being greater than a preset information similarity threshold; determining preset problems whose problem similarity to the target problem is greater than the preset problem similarity threshold, and determining the knowledge information corresponding to the determined preset problems as the basis information for answering the target problem.
Need to check novelty before this filing date? Find Prior Art