面向情报立方体的战略态势解析方法研究

As China’s increasingly complex security situation and expanded national interests, it is required that intelligence should be able to analyze the global strategic situation effectively and promptly. Due to the big data properties of intelligence, the lack of effective and efficient data processing results in the low level of data integration and the weak intelligent analysis, which largely reduce the utilization of intelligence resources, the ability of the comprehensive intelligence judgments, early forecasting and early warning. According to the requirements of the intelligence unified understanding, intelligent association and analysis of the cross-regional, cross-cultural and cross-spatial global strategic situation, this project will propose the concept of "intelligence Cube" by fusing the attributes, relationships, behaviors, events, rules and predictions. The four core research problems of this project are as follows: 1) the cross-regional strategic situation data analysis and understanding; 2) the cross-cultural strategic situation data analysis and understanding; 3) the cross-spatial strategic situation data analysis and understanding;4)the deep understanding of the global strategic situation. These problems will 1) transform the complex intelligence data to be simple to understand, 2) extract intelligence and knowledge for the intelligent analysis and unified understanding to intelligence in an efficient, accurate and comprehensive way, 3) generate the cross-time and cross-space strategic intelligence cube with complete properties,.clear relations, 4) build intelligence object space and support the intelligence analysis, such as the trend analysis, abnormal detection, early crisis warning. Finally, this project will provide theoretical methods and technical reserves for the dynamic monitoring of the global strategic situation, early warning and emergency treatment.

我国日趋复杂的安全形势和利益拓展，要求情报部门能够及时有效地开展全球战略态势分析。当前情报信息呈现大数据特征，而融合水平低、智能分析弱的技术瓶颈导致处理能力严重不足，影响了情报信息资源利用效率、情报综合研判和预测预警。本项目针对上述问题，根据跨区域、跨文化和跨空间背景下全球战略态势情报统一理解、融合关联、智能分析的需求，拟提出包含属性、关系、行为、事件、规则以及趋势的“情报立方体”概念，重点研究1）跨区域战略态势数据的计算与理解；2）跨文化战略态势感知数据的计算与理解；3）跨空间战略态势数据的计算与理解；4）全球战略态势深度研判技术。从而高效、准确、全面地提炼出满足情报智能分析、统一理解利用的情报和知识，生成属性完备、关系清晰、跨越时空的战略情报立方体，并进一步构建情报对象空间，支持异常发现、趋势分析和危机预警等业务，为全球战略态势的动态监控、预警预报与应急处理提供理论方法和技术储备。

本项目基于情报分析由人工经验型向智能分析型转变的迫切需求。面向当前情报数据融合水平低、智能分析弱的技术瓶颈导致情报大数据处理能力严重不足，情报分析以人工定性为主，导致情报信息资源利用效率低、情报综合研判和预测预警能力弱的现状。本项目提出面向情报立方体的战略态势解析方法研究。根据跨区域、跨文化和跨空间背景下全球战略态势情报统一理解、融合关联、智能分析的需求，开展情报立方体构建和面向情报立方体的全球战略态势解析方法研究，将为融合情报大数据，高效、准确、全面地构建属性完备、关系丰富、跨越时空的战略情报对象空间提供技术手段，从而为国际关系、社会危机、军事行动等方面的情报研判与预测提供基础支撑。.1..面向现实空间、赛博空间以及二元空间联合感知，开展了深入的模型和算法研究，包括：跨传感器区域分析、目标实体建模和交互分析、社会群体发掘与关系分析、情境上下文感知建模、排序预测模型和算法、实体行为和意图分析、事件态势分析与预测等；.2..完成了现实空间和赛博空间感知数据集，不实信息数据集和全球恐怖事件数据集等基础性数据集，有效支撑了理论研究和应用示范工作的开展；.3..建设了面向情报立方体的战略态势解析示范系统，核心包括：全球恐怖事件序列预测软件、多媒体危机事件综合评估软件和全球媒体态势感知平台；.4..发表学术论文41篇；出版英文专著1本；撰写国防专利3项；申请/授权国家发明专利22项；培养博士8名、硕士6名；参加国际学术会议14人次。