A knowledge graph-based financial information management method

By constructing a knowledge graph-based financial information management system, the challenges of heterogeneity of multi-source financial data and risk identification were solved, enabling efficient and unified processing and intelligent management of financial data, and improving the accuracy of risk identification and the real-time nature of auditing.

CN121456043BActive Publication Date: 2026-05-26BEIJING SHANGXIN TECHNOLOGY CO LTD
View PDF 2 Cites 0 Cited by

Patent Information

Authority / Receiving Office
CN · China
Patent Type
Patents(China)
Current Assignee / Owner
BEIJING SHANGXIN TECHNOLOGY CO LTD
Filing Date
2025-11-05
Publication Date
2026-05-26

AI Technical Summary

Technical Problem

Existing financial management platforms suffer from heterogeneity, delays, and a lack of semantic understanding capabilities in integrating multi-source financial data and identifying risks, making it difficult to meet the panoramic traceability requirements in financial auditing and regulatory scenarios.

Method used

By adopting a knowledge graph-based approach, an initial financial knowledge graph is constructed, embedding time, business, risk, and responsibility tags. A graph update scheduling mechanism is set up, an improved NodeFormer model is used for risk path identification, and a graph semantic question-and-answer structure and visualization module are constructed to achieve unified data processing, risk identification, and intelligent query.

Benefits of technology

It achieves a unified structured representation of multi-source financial data, timely dynamic graphs, and high accuracy in risk identification, thereby improving the level of intelligence in financial information management and audit support capabilities.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN121456043B_ABST
    Figure CN121456043B_ABST
Patent Text Reader

Abstract

This invention discloses a knowledge graph-based financial information management method, comprising the following steps: S1, collecting multi-source heterogeneous financial data to generate an initial financial dataset; S2, extracting financial entities and semantic relationships, constructing a set of triples to form an initial financial knowledge graph; S3, performing semantic enhancement processing to construct a semantically enhanced financial graph; S4, setting a graph update scheduling mechanism, performing dynamic update operations, and generating a dynamic graph structure; S5, inputting the graph into an improved NodeFormer model and outputting risk path identification results; S6, constructing a graph semantic question-and-answer structure and outputting structured query response results; S7, constructing a graph visualization module to generate graph analysis views and audit suggestion results. This invention improves the accuracy and tracking capability of risk paths in financial data, effectively enhancing the transparency, response efficiency, and risk control capabilities of the financial audit process.
Need to check novelty before this filing date? Find Prior Art