摘要
突发事件的爆发伴随着大量相关信息涌现,突发事件大数据虽然丰富了相关信息类型和规模,但造成了应急决策的信息利用障碍。面向应急决策全流程的信息需求进行突发事件大数据信息融合是提升突发事件大数据的应急决策支持力的有效途径。本文以构建面向应急决策全流程信息需求的大数据融合框架及融合路径为目的,首先依据应急决策流程从信息内容、信息特征两个方面归纳突发事件应急决策信息需求,与大数据环境下可获得的应急信息进行对比,识别出应急信息融合的具体目标及问题;进而,构建以应急数据模型为桥梁、面向应急决策服务的信息融合服务整体框架;最后,从数据层、语义层和服务层逐层剖析应急信息融合中的具体问题,并提出相应的信息融合实现路径。研究结论有助于指导突发事件大数据融合系统构建,也为面向决策的信息分析提供实践参考依据。
The outbreak of emergent events is always accompanied with a large number of relevant information.Big data of emergent events not only enriches the type and scale of emergency information,but also hinders the information utilization of emergency decision-making.Fusing emergency information oriented to the information needs throughout the decision-making process is an effective approach to enhance the decision-making support of big data.This study aims to build an emergency big data integration framework and explore the detailed information fusion methods.First of all,we develop information demands both from information contents and characteristic according to the decision-making process of emergent events.Then,identifying the objectives and problems of information fusion with a comparison between information needs of emergent event decision-making and the available emergency information in big data environment.Accordingly,a service-oriented framework of information fusion for emergent event decision-making has been developed with a consideration of data models as the bridge.Finally,specific issues and realizing paths of information fusion have been put forward according to the three layers of emergency information fusion,which include data layer,semantic layer and service-oriented layer.The research findings will help construct emergency integration system of big data,and yield practical implications to decision-oriented information analytics.
出处
《图书情报知识》
CSSCI
北大核心
2018年第5期95-104,共10页
Documentation,Information & Knowledge
基金
国家自然科学基金重大课题"国家安全大数据综合信息集成与分析方法"(71790612)的成果之一
关键词
大数据
应急决策流程
应急信息需求
细粒度信息融合
语义融合
Big data
Emergent event decision-making process
Emergency information needs
Fine-grained information fusion
Semantic fusion