A Public Safety Risk Assessment and Early Warning Method Integrating Multi-Source Heterogeneous Data
By combining graph attention networks and knowledge graphs, the public safety risk assessment method is optimized, which solves the problem of insufficient processing capabilities for real-time dynamic data and achieves more efficient and interpretable risk assessment.
Patent Information
- Application Number
- CN202610207775.9
- Authority / Receiving Office
- CN · China
- Patent Type
- Patents(China)
- Current Assignee / Owner
- Filing Date
- 2026-02-12
- Publication Date
- 2026-05-26
- Estimated Expiration
- 2046-02-12
AI Technical Summary
Existing technologies are insufficient in processing real-time dynamic data in public safety risk assessments, especially video and text data. They also rely on rules defined by expert experience, resulting in low efficiency and interpretability of risk assessments.
By combining graph attention networks and knowledge graphs, features of multi-source heterogeneous data are extracted and mapped onto knowledge graph nodes. Graph attention networks are used to optimize inference, generating representation vectors that contain the global graph structure context. Probabilistic inference is then performed through vectorized representations of energy functions and logical rules to output risk assessment results.
It improves the efficiency and interpretability of public safety risk assessment, reduces reliance on manually defined rules, enables a better understanding of complex risk scenarios, and generates interpretable reasoning paths and risk assessment conclusions.
Smart Images

Figure CN121684662B_ABST
Abstract
Citation Information
Patent Citations
Urban rail transit comprehensive monitoring method and system based on multi-source heterogeneous data
CN121388760A