基于特征构建的多源数据与异构关系融合的资源分配方法
By constructing a resource allocation method that integrates multi-source data and heterogeneous relationships, the problem of rigid resource allocation strategies in existing technologies is solved, achieving efficient and accurate resource allocation, adapting to changes in external factors, and providing auditable solutions.
Patent Information
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- ANHUI UNIV
- Filing Date
- 2026-06-16
- Publication Date
- 2026-07-17
AI Technical Summary
Existing resource allocation methods cannot effectively capture and utilize the complex relationships between entities, resulting in rigid allocation strategies that are difficult to balance fairness and efficiency, and lack the ability to dynamically quantify the strength, type, and timeliness of relationships.
By constructing a resource allocation method based on feature construction and fusion of multi-source data and heterogeneous relationships, the method includes acquiring numerical and textual data, constructing node features, generating heterogeneous graphs, encoding relationship subgraphs using graph neural networks, and fusing representations through a relationship attention mechanism, ultimately optimizing resource allocation under constraints.
It achieves more efficient and accurate resource allocation, can adapt to changes in external factors, takes into account both policy guidance and actual needs, and provides auditable allocation schemes.
Smart Images
Figure CN122414741A_ABST