Construction and retrieval enhancement method and device based on medical insurance knowledge graph

By constructing a multi-dimensional segmented medical insurance knowledge graph and combining it with intent recognition and timeliness filtering, the problems of data accuracy and timeliness in medical insurance knowledge graph retrieval are solved, and efficient and accurate retrieval results are generated.

CN121833970BActive Publication Date: 2026-07-24CAPINFO CO LTD
View PDF 2 Cites 0 Cited by

Patent Information

Authority / Receiving Office
CN · China
Patent Type
Patents(China)
Current Assignee / Owner
CAPINFO CO LTD
Filing Date
2025-12-30
Publication Date
2026-07-24

AI Technical Summary

Technical Problem

Existing medical insurance knowledge graph retrieval methods have low accuracy in search results, suffer from semantic noise and low signal-to-noise ratio, and are particularly inefficient when dealing with complex cross-domain queries and high-frequency hot topics. Furthermore, they are difficult to effectively handle the timeliness and version control of knowledge.

Method used

By constructing a multi-dimensional medical insurance knowledge graph, the graph is logically divided into independent subgraphs such as policy, resources, and medicine. In conjunction with intent recognition, routing retrieval is performed, query decomposition and hotspot shortcut mechanisms are introduced, and filtering is performed using time windows and time parameters to achieve parallel retrieval and the construction of shortcut edges, ensuring the accuracy and timeliness of retrieval results.

Benefits of technology

It significantly improves the accuracy and relevance of search results, solves the performance bottleneck of complex queries, ensures the timeliness and accuracy of information, and achieves a significant improvement in the search signal-to-noise ratio and a faster system response speed.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN121833970B_ABST
    Figure CN121833970B_ABST
Patent Text Reader

Abstract

The application provides a construction and retrieval enhancement method and device based on a medical insurance knowledge graph, and relates to the field of retrieval. The method: establishes cross-domain anchor nodes in a graph database, takes the target relationship between the pharmacological properties of the target drug chemical components, the target medical insurance reimbursement rules and the target designated medical institutions as the edges corresponding to the connection of the anchor nodes, sets the domain label corresponding to the target relationship on the edge, adds the time window on the attribute of the edge according to the pharmacological properties of the target drug chemical components and the target medical insurance reimbursement rules, and obtains the medical insurance knowledge graph constructed based on the domain subgraph; analyzes the query request by using a classification model and identifies multiple user intents of the query request, matches the user intents with the domain labels, and obtains multiple target domain labels matched with the user intents; and according to the target domain labels, the query request is disassembled into atomized query data for different domain corresponding target domain subgraphs.
Need to check novelty before this filing date? Find Prior Art