一种移动应用智能代理方法
By constructing a long-term experience base and a multi-agent collaborative decision-making framework, and by utilizing visualization to explore historical and multimodal information fusion, the problems of unreasonable inference task allocation and insufficient cross-scenario generalization ability in intelligent agent methods for mobile applications are solved, realizing intelligent task processing and multimodal information utilization across scenarios.
Patent Information
- Authority / Receiving Office
- CN · China
- Patent Type
- Patents(China)
- Current Assignee / Owner
- BEIJING UNIV OF POSTS & TELECOMM
- Filing Date
- 2025-09-09
- Publication Date
- 2026-07-17
AI Technical Summary
Existing intelligent agent methods for mobile applications fail to effectively utilize multimodal information and lack cross-scenario generalization capabilities, resulting in unreasonable allocation of inference tasks and limitations in experience reuse, and the multimodal capabilities are not fully explored.
We construct a long-term experience base and a multi-agent collaborative decision-making framework. By visually exploring the fusion of historical and multimodal information, we achieve cross-scenario experience sharing and task planning. The multi-agent collaborative decision-making framework includes modules for task planning, experience extraction, action decision-making, execution, and result inspection. We also utilize visual information to form a knowledge graph.
It enables intelligent task processing across scenarios, improves the rationality of task allocation and the utilization of multimodal information, and enhances the accuracy and efficiency of decision-making.
Smart Images
Figure CN121125830B_ABST