一种基于多智能体协同的不完整数据趋势问答方法

By employing a multi-agent collaborative mechanism and a lightweight meta-controller based on reinforcement learning, the challenges of sub-table location and missing value repair caused by missing table metadata and time-series data were solved, enabling efficient and accurate trend analysis and improving the accuracy and robustness of analysis in complex scenarios.

CN121958360BActive Publication Date: 2026-07-17ZHEJIANG UNIV
View PDF 2 Cites 0 Cited by

Patent Information

Authority / Receiving Office
CN · China
Patent Type
Patents(China)
Current Assignee / Owner
ZHEJIANG UNIV
Filing Date
2026-04-02
Publication Date
2026-07-17

Smart Images

  • Figure CN121958360B_ABST
    Figure CN121958360B_ABST
Patent Text Reader

Abstract

本发明公开了一种基于多智能体协同的不完整数据趋势问答方法，包括：获取自然语言查询及存在缺失的原始表格数据；构建包含提取器和修正器的序贯博弈模型，利用表格内部关联信息交互推理，定位目标子表；针对时序数据缺失，构建包含行填充器和列填充器的不完全信息博弈模型，分别生成基于属性关联视图及时序趋势视图的填补子表；分别执行趋势计算后，由协调器对填补结果及趋势结论进行综合决策并输出。本发明引入轻量级元控制器及分阶段训练策略，在无外源知识辅助情况下，有效解决了子表定位困难及缺失数据填补不准的问题，提升了面向不完整数据趋势问答的鲁棒性。
Need to check novelty before this filing date? Find Prior Art