基于社区搜索的面向联邦数据湖仓的客户运营方法及系统
By constructing a local multimodal heterogeneous graph and utilizing modality-aware feature encoding and secure anchor mapping, the problems of low customer group identification accuracy and insufficient privacy protection in the federated data lake warehouse environment are solved, enabling accurate and secure cross-domain customer operations.
Patent Information
- Authority / Receiving Office
- CN · China
- Patent Type
- Patents(China)
- Current Assignee / Owner
- NORTHEASTERN UNIV CHINA
- Filing Date
- 2025-10-30
- Publication Date
- 2026-07-17
AI Technical Summary
Existing technologies cannot effectively integrate multimodal data in a federated data lake warehouse environment, lack community structure awareness, resulting in low customer group identification accuracy, insufficient privacy protection, loose cross-source community structure, poor cross-domain operation performance, and the risk of privacy leakage.
By constructing a local multimodal heterogeneous graph in a federated data lake warehouse environment, extracting features using a modality-aware feature encoder, fusing features using a multi-head attention mechanism, and constructing anchor mappings by combining homomorphic encryption and secure multi-party computation, secure cross-source community search is achieved, generating structure-aware embeddings, and identifying target customer groups through a community search algorithm.
It enables precise identification of customer groups, improves cross-domain operation effectiveness, enhances the operability and explainability of operation strategies, and supports refined customer segmentation and cross-regional marketing, all while ensuring privacy compliance.
Smart Images
Figure CN121434446B_ABST