Microblog-oirented Hot Topics Discovery and Tracking
|School||South China University of Technology|
|Keywords||Microblog Hot Topic Discovery Topic Tracking Single-Pass SemanticFramework|
With the continuous development of web2.0, the applications based on web2.0,which makes a big change of the netizens way of communication, appear constantly.Microblog is recently got the rapid development based on the development of web2.0.Internet users can use Microblog to express their own information, pay attention to theother information or forward, comment on other people’s information. But in this way,Internet users will easy to fall into local information and ignore the overallinformation. So, this paper using the information of netizens research on theMicroblog hot topics. The main works in this paper are as follows:1. This paper analyzes the construction of the Microblog and gives the way toextract the information of Microblog. Because the traditional extraction in Microblogis limit appeared, so this paper proposed Microblog crawler based on Ajax. At thesame time, we discussed different noise of Microblog information, and the differentsituation to filter different noise.2. From the reason of the length of Microblog, which is limited by140words, thetraditional way of topic discovery is fail. In the way, we give the concept ofMicroblog Discussing Tree and the algorithm to merger Microblog.3. To research on the topic of Microblog, we give the concept of MicroblogSemantic Framework. But there are certain defect if use the semantic Framework onlyor Single-Pass method only on topic discovery. So an improved method for discoveryhot topic is proposed based on merger Single-Pass Method and Microblog SemanticFramework. Experimental results show that the method can improve the accuracy tosome extent.4. In order to track the topic of Microblog, we define the value of Microblog,value of Microblog Discussing Tree and value of hot topic. The method of calculatingtopic the current energy is also discussed.