信息网络安全 ›› 2015, Vol. 15 ›› Issue (1): 16-23.doi: 10.3969/j.issn.1671-1122.2015.01.004

• 技术研究 • 上一篇    下一篇

基于微博的安全事件实时监测框架研究

李凌云1(), 敖吉2, 乔治3, 李剑1   

  1. 1.北京邮电大学计算机学院,北京100876
    2.中国科学院信息工程研究所,北京 100093
    3.中国科学院计算技术研究所,北京 100190
  • 收稿日期:2014-12-09 出版日期:2015-01-10 发布日期:2015-07-05
  • 作者简介:

    作者简介: 李凌云(1991-),男,河南,硕士研究生,主要研究方向:信息安全与数据挖掘;敖吉(1988-),女,内蒙古,硕士研究生,主要研究方向:面向社会媒体数据流的安全态势分析;乔治(1986-),男,山东,博士研究生,主要研究方向:数据挖掘与社会计算;李剑(1976-),男,陕西,副教授,博士,主要研究方向:信息安全和数据挖掘。

  • 基金资助:
    国家自然科学基金[61472048,61402058]

Research on Security Event Real-time Monitoring Framework Based on Micro-blog

LI Ling-yun1(), AO Ji2, QIAO Zhi3, LI Jian1   

  1. 1. School of Computer Science, Beijing University of Posts and Telecommunications, Beijing 100876, China
    2. Institute of Information Engineering, CAS, Beijing 100093, China
    3. Institute of Computing Technology, CAS, Beijing 100190, China
  • Received:2014-12-09 Online:2015-01-10 Published:2015-07-05

摘要:

文章根据微博事件发展规律和传播特点,在微博社会感知器网络基础上,提出了针对微博安全事件的实时监测框架,该框架包含若干项核心算法,如异常检测算法、地理位置定位算法、相关事件推荐算法和事件相关度分析算法。基于此框架,文章实现了微博事件实时监测系统。该系统采用混合网络爬虫和开放API接口方式采集微博数据,并实现了事件检索模块、事件实时监测模块和热点模块。同时该系统以多维度展示微博事件结果信息,且运行稳定、效果良好。总体上看,文章主要解决的问题是探索虚拟社交网络与物理世界时空相关性,监测特定事件,并在其爆发前发现并进行地理定位,从而提供预警。

关键词: 微博事件, 实时监测, 异常检测, 地理定位

Abstract:

According to the discipline of event’s development and the social characteristic of event’s propagation, this paper proposes a framework of real-time monitoring events which propagating on micro-blog, based on the theory of social sensor network, and this framework includes several key algorithms, such as abnormal detection algorithm, geography location positioning algorithm, related events recommendation algorithm, and event correlation analysis algorithm. Based on this framework, this paper develops and implements a real-time monitoring system about micro-blog events. This system applies hybrid web crawler and the way of open API interface to capture micro-blog data, and also implements the event retrieval module, real-time monitoring module and hot topic module. This system also displays the result information of micro-blog event in multiple dimensions, and operates stably. In conclusion, this paper is to explore the field of spatial-temporal correlation between the virtual network and the physical world, monitor the specific "event", and position its location before outbreak, and provide early warning.

Key words: micro-blog events, real-time monitoring, anomaly detection, location

中图分类号: