一种多源数据融合的企业征信报告智能生成方法和产品
By combining graph neural network models and the MCP protocol, entity alignment and conflict resolution of multi-source data are achieved, generating a unified corporate credit report. This solves the problems of system coupling and logical contradictions in multi-source data fusion, improves the accuracy and auditability of the report, and is applicable to corporate background checks, credit approval, and supply chain risk assessment.
Patent Information
- Authority / Receiving Office
- CN · China
- Patent Type
- Patents(China)
- Current Assignee / Owner
- 知呱呱(天津)大数据技术有限公司
- Filing Date
- 2026-04-28
- Publication Date
- 2026-07-17
AI Technical Summary
Existing technologies for generating corporate credit reports suffer from bottlenecks in the coupling and expansion of multi-source data access systems, a lack of deep semantic-level fusion of multi-source heterogeneous data, and a lack of natural language interaction and uncontrollable generation, leading to logical contradictions and factual errors in report generation.
Entity alignment is performed using a graph neural network model, a credibility weighting mechanism based on time decay factor is introduced for conflict resolution, and the MCP protocol is used to implement standardized tool encapsulation and dynamic scheduling of multi-source data to construct a business association graph. Reports are generated by combining the context region and constraint instructions of the large model.
It achieves a unified model for multi-source data, improving data consistency and accuracy. The generated credit reports, while ensuring the flexibility of natural language interaction, meet commercial compliance-level accuracy and auditability, and are suitable for corporate background checks, credit approval, and supply chain risk assessment.
Smart Images
Figure CN122113870B_ABST