Approach to extracting hot topics based on network traffic content

作     者：Yadong ZHOU Xiaohong GUAN Qindong SUN Wei LI Jing TAO 

作者机构：MOE Key Lab for Intelligent Networks and Network SecurityState Key Lab for Manufacturing SystemsXi’an Jiaotong UniversityXi’an 710049China Department of AutomationTsinghua National Lab for Information Science and TechnologyTsinghua UniversityBeijing 100084China School of Computer Science and EngineeringXi’an University of TechnologyXi’an 710048China 

出 版 物：《Frontiers of Electrical and Electronic Engineering in China》 (中国电气与电子工程前沿（英文版）)

年 卷 期：2009年第4卷第1期

页      面：20-23页

学科分类：08[工学] 0812[工学-计算机科学与技术（可授工学、理学学位）] 

基　　金：was supported by the National Natural Science Foundation of China (Grant No.60574087) the Hi-Tech Research and Development Program of China (2007AA01Z475,2007AA01Z480,2007A-A01Z464) the 111 International Collaboration Program of China 

主　　题：hot topic extraction network traffic content Internet public opinion analysis 

摘      要：This article presents the formal definition and description of popular topics on the Internet,analyzes the relationship between popular words and topics,and finally introduces a method that uses statistics and correlation of the popular words in traffic content and network flow characteristics as input for extracting popular topics on the *** on this,this article adapts a clustering algorithm to extract popular topics and gives formalized *** test results show that this method has an accuracy of 16.7%in extracting popular topics on the *** with web mining and topic detection and tracking(TDT),it can provide a more suitable data source for effective recovery of Internet public opinions.