大数据背景下网络突发事件动态监测研究

发布时间：2018-02-25 20:13

本文关键词： 大数据网络突发事件动态监测 Kleinberg算法　出处：《统计研究》2017年02期 　论文类型：期刊论文

【摘要】：通过互联网以及社交平台的数据监测社会突发异常事件是当前社交网络传播研究的热点问题,本文基于大数据背景,对网络突发事件动态监测算法进行了创新性研究。论文首先介绍了常用的几种网络动态监测算法,并指出了其在社交网络中应用的局限性,提出了基于Kleinberg的改进算法,最后用改进的算法进行了实证研究,研究结果显示:改进后的算法通过给定合理的调节参数,能迅速准确地监测到网络异常状况的发生,从而不仅避免了传统主观上直接通过突发事件出现的频次来划定临界值的不科学性和直接使用微博量排行榜方式所产生的时间上的滞后性,也避免了直接通过划定一个固定的微博增加量来判断是否有异常事件发生的不合理性。本文在大数据背景下为网络突变事件的动态监测提供了新的研究方法,为政府加强网络安全建设、优化网络监管模式、净化互联网环境具有一定的实践指导意义。
[Abstract]:Monitoring social outbursts through the Internet and social platform data is a hot issue in the research of social network communication. This paper is based on big data background. This paper first introduces several commonly used network dynamic monitoring algorithms, points out the limitations of their application in social networks, and puts forward an improved algorithm based on Kleinberg. Finally, an empirical study with the improved algorithm is carried out. The results show that the improved algorithm can quickly and accurately monitor the abnormal situation of the network by giving reasonable adjustment parameters. This not only avoids the unscientific nature of the traditional subjective determination of critical values directly through the frequency of unexpected events and the lag in time caused by the direct use of Weibo's ranking method. It also avoids the irrationality of judging whether there are abnormal events directly by defining a fixed increase of Weibo. This paper provides a new research method for dynamic monitoring of network abrupt events under the background of big data. It is of practical significance for the government to strengthen the construction of network security, optimize the mode of network supervision, and purify the Internet environment.
【作者单位】：对外经济贸易大学统计学院;中华女子学院管理学院;
【分类号】：C812

【相似文献】