项目名称: 面向复杂情报的大数据分析方法与决策支持
项目编号: No.U1435220
项目类型: 联合基金项目
立项/批准年度: 2015
项目学科: 其他
项目作者: 胡晓惠
作者单位: 中国科学院软件研究所
项目金额: 500万元
中文摘要: 本项目针对复杂情报大数据处理、分析与决策的挑战性问题,在对相关的理论基础与方法进行深入调研与论证的基础上,重点研究大数据分析方法,包括基于情报逻辑语义的数据表示、压缩感知的稀疏处理、融合反馈的认知方法。重点突破复杂情报数据的特征提取与融合、基于主动规则的异常检测、基于时空线索的突发事件预警、潜在信息关联关系挖掘、复杂多维情报数据动态视图生成等技术,在大数据分析应用开发与运行原型系统基础上部署典型应用。本项目研究成果对于提高与国家安全相关的特定领域内情报获取的准确度、应对策略的前瞻性、合理性和科学性上具有重要意义,可进一步丰富和发展我国情报大数据分析的理论方法和技术体系,提升我国信息化建设综合实力、在军民共性技术领域发挥重要作用。
中文关键词: 大数据;情报数据;潜在信息关联挖掘;异常检测;决策支持
英文摘要: The increasing complexity of information-data brings great challenges for the corresponding data processing, analysis and decision-making process. Facing the challenges, we will mainly investigate the methodologies for complicated information-data analysis, including the semantic representation of data, compressive sensing oriented sparse processing, and fusion feedback based recognizing methods. Furthermore, we will be devoted to the research of the following key techniques, involving characteristics extraction and fusion from complicated data, active rule based abnormal detection, spatio-temporal oriented emergence sensing, relation mining of potential information, and dynamic view generation for multi-dimensional data. With the prototype implemented with the above techniques, we will validate the effectiveness through classical case studies. The achievements will not only increase information-data acquiring accuracy, strategy predictability and reasonability for the related areas, but also enrich the theoretical approaches and technical frameworks of information processing and analysis, which are essential to enhance the comprehensive strength for both military and civilian areas.
英文关键词: Big data;Information-data;Potential information relation mining;Anomaly detection;Decision support