多维度跨领域大数据智能挖掘分析是当前大数据挖掘研究和应用发展的重要方向，问题任务的复杂性、信息知识的不确定性和数据的动态变化，是其中的关键挑战。本申请项目拟借鉴人类多粒度任务分解、信息选择遗忘和联想记忆机理，探索研究大数据复杂问题任务的多粒度分解与联合问题求解机制，主要包括：基于大数据的不确定性信息知识的多粒度知识表示与处理；基于分解机制的多维度跨领域大数据中的自适应多粒度知识发现；基于概念内涵与外延之间双向认知变换的信息选择遗忘和联想记忆模型；融合选择遗忘和联想记忆机理的多粒度知识动态演化模型；大数据复杂任务的多粒度分解与联合问题求解机制；并在生态环境智能化预测预警、工业系统智能优化调度管理等领域应用系统中进行实验验证。相关研究成果将形成一套有效的多粒度分解与联合问题求解模型理论与方法。

Intelligent analysis and mining of multi-dimensional and cross-domain big data is an important research and application direction of big data processing. The complexity of a problem and task, the uncertainty of information and knowledge, and the dynamic evolution of data, are its key challenges. This project will explore the multi-granularity decomposition and joint problem solving mechanism of big data complex tasks, using such human being intelligence mechanisms as multi-granularity task decomposition, selective information forgetting and associative memorizing. The major research tasks include: multi-granularity knowledge representation and processing of uncertain information and knowledge of big data; adaptive multi-granularity knowledge discovery from multi-dimensional and cross-domain big data based on task decomposition mechanism; selective information forgetting and associative memorizing model based on the bidirectional cognitive transformations between concept’s extension and intension; multi-granularity knowledge dynamical evolution model based on the selective forgetting and associative memorizing mechanisms; multi-granularity decomposition and joint problem solving mechanism of big data complex tasks. These theoretical research results will be used to solve the real life problems such as water quality prediction and early warning, intelligent optimization and dispatching management of industrial systems. A multi-granularity decomposition and joint problem solving model of big data complex tasks will be developed at last.