A knowledge graph retrieval enhancement generation method adaptive to user query complexity
CN120687577BActive Publication Date: 2026-08-28CHINA UNIV OF MINING & TECH
Patent Information
- Application Number
- CN202510953714.2
- Authority / Receiving Office
- CN · China
- Patent Type
- Patents(China)
- Current Assignee / Owner
- Filing Date
- 2025-07-10
- Publication Date
- 2026-08-28
- Estimated Expiration
- 2045-07-10
AI Technical Summary
Technical Problem
现有知识图谱检索增强生成方法仍然存在知识检索范围和检索策略的灵活性不足的缺点，即现有知识图谱检索增强生成方法通常利用同一知识检索策略在固定检索范围内处理所有用户查询，而没有考虑用户查询复杂程度的影响，具体来说：(1)对于简单查询，只需在小范围内识别少量最相关的目标知识，其检索范围小，目标知识少，检索策略简单；(2)对于复杂查询，则需要在更大范围内识别出更多数量的目标知识，其检索范围大，目标知识多，检索策略复杂
Benefits of technology
[0087]有益效果：相对于现有技术通常采用同一策略处理所有用户查询，导致知识检索范围和检索策略的灵活性不足，本发明提供的用户查询复杂度自适应的知识图谱检索增强生成方法，具有如下优势：1、考虑到用户查询复杂程度的影响，本发明可以根据查询复杂程度自适应地调整知识检索范围和检索策略，增强模型的灵活性；2、对于简单用户查询，所需目标知识较少，可以适当缩紧检索范围和检索策略，以保证检索精度并加快检索效率；3、对于复杂用户查询，所需目标知识较多，可以适当放松检索范围和检索策略，以保证检索效率并提高检索精度；4、本发明打破了现有知识检索策略的静态限制，实现了根据用户查询复杂程度灵活确定知识检索范围和检索策略，达到了兼顾检索精度与检索效率的目的。
✦ Generated by Eureka AI based on patent content.
Smart Images
Figure CN120687577B_ABST
Abstract
The application discloses a knowledge graph retrieval enhancement generation method adaptive to user query complexity, which quantifies user query complexity according to complexity measurement indexes contained in the user query, and analyzes and completes the hidden logic of the user query; then, a corresponding relationship between the user query complexity and the knowledge graph retrieval range and the knowledge graph retrieval strategy is established, and the knowledge graph retrieval range is adaptively adjusted in combination with the user query complexity; next, an optimal set of knowledge reasoning paths is adaptively screened out in the determined knowledge graph retrieval range based on reinforcement learning; finally, a large language model is used to answer the user query based on the selected knowledge reasoning paths. The application breaks the static limitation of the existing knowledge retrieval strategy, can flexibly determine the knowledge retrieval range and the retrieval strategy according to the complexity of the user query, and takes into account the retrieval accuracy and the retrieval efficiency.
Need to check novelty before this filing date? Find Prior Art
Citation Information
Patent Citations
Aviation standard question and answer optimization method and system based on atlas and document data
CN119621894A
Multi-hop search for entity relationships
US20220164679A1