The Analysis Of Microblogging Public Opinion Based On The Hadoop

Given the propagation characteristics of micro-Bo is not only the media, there are some typical characteristics of social networks, so the study of microblogging can be done from various angles social networks, news dissemination. Twitter is an online social network, we design for the microblogging Web crawler to crawl data. Crawl on the basis of the data on the microblogging network conducted a preliminary analysis, and proposed a new evaluation algorithm microblogging network node influence.The main work of this paper the following three aspects:First, the design of a common reptile microblogging, microblogging can be a lot of data to crawl and storage. Secondly, through the microblogging network structure analysis, we found that the microblogging network has fat tail characteristics and small world, but also of the microblogging network obeys a power law distribution. It is these features are typical characteristics of complex networks. Microblogging feature has a complex network, based on this paper, the method of analysis of complex networks microblogging analyzed. Thirdly, over the Internet search engine PageRank algorithm thinking we propose a new relationship importance algorithms (IIR, Important-Relationship Rank) to evaluate the influence of microblogging network nodes. The algorithm takes into account users microblogging network level of activity and influence. This article uses the Hadoop distributed computing platform, through the MapReduce programming framework implements this algorithm, using distributed computing to reduce large-scale data analysis to calculate the amount of time, to improve the efficiency of the analysis. By calculating the influence of the size of each network node to identify the node play a key role in the network information dissemination process, which is the study of microblogging network dissemination of information is helpful, but also for the microblogging network monitoring public opinion offer help.Microblogging as a new platform for the dissemination of information in the application and research have a very bright future, especially recommended information, news dissemination, user behavior analysis and other aspects of human dynamics. Microblogging public opinion research purpose is to discover its potential value in the application and research, which is driving force behind advances in technology.