Social governance element risk assessment method based on dynamic gating mechanism
By processing heterogeneous graph data of social governance through dynamic gating and relational attention mechanisms, the problem of low accuracy in risk prediction caused by imbalance in node characteristics is solved, thereby improving the accuracy and robustness of node risk assessment.
Patent Information
- Application Number
- CN202511777012.X
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2025-11-28
- Publication Date
- 2026-03-06
AI Technical Summary
In social governance scenarios, the imbalance in the distribution of node type characteristics leads to low accuracy in risk prediction, and the problems of noise propagation and representation bias are difficult to solve.
By employing a dynamic gating mechanism, node feature processing and unified projection are performed on heterogeneous graph datasets to generate dynamic sparse activation weights and attention aggregation results. Combined with a relational attention mechanism, a risk prediction model is trained to achieve node risk assessment.
It effectively alleviates the problems of noise propagation and representation bias caused by imbalanced node features, improves the classification accuracy and robustness of multi-type nodes, is suitable for heterogeneous graph scenarios with imbalanced features, and can adaptively adjust neighbor weights and enhance the representation ability of weak semantic nodes.
Smart Images

Figure CN121616094A_ABST