Influence Analysis For Internet Public Opinions In Microblogs

With the rapid development of Internet, the number of Sina weibo user is growing at a very high speed. Micro-blog is a kind of social networking platform that allows users to publish short text messages and share information by attention mechanism. In recent years, micro-blog with its vast amounts of real-time information, open, low threshold and many other characteristics is now becoming a fermentation place of popular event. The study of its user’s influence has great practical significance in the increasingly fierce social contradictions today. Micro-blog’s attention mechanism and information’s one-way circulation makes micro-blog not only has the social attribute,but also an important medium of information dissemination. In this paper, through the deep analysis of the characteristics of micro-blog’s complex networks and combining with the micro-blog network information dissemination characteristics, study Micro-blog network node’s influence, and eventually achieved good experimental results.The main work of this paper are as follows:1) We analyze micro-blog’s media properties and propagation mode, based on the analysis of complex network characteristics, it proved the feasibility of using the social network’s knowledge to analyze micro-blog network node. 2)Based on social network node properties and micro-blog propagation analysis, defines the micro-blog network node influence,combined with PageRank algorithm, put forward the micro-blog user’s influence algorithm which we use both the following relationship network and the micro-blog’s information dissemination network. 3) We use sina weibo’s apis and third party application Oauth certification and distributed crawler to get huge amounts of sina weibo data, in Hadoop distributed system, we use MapReduce to achieve user’s impact evaluation algorithm. 4)On the hadoop distributed system, massive micro-blog data is used to calculate micro-blog user’s influence, the experimental results and the actual micro-blog data were analyzed together. Experiments show that, this evaluation method can effectively reflect the micro-blog user’s real influence in micro-blog service, it is suitable for measuring the micro-blog user’s influence.Micro-blog network contains rich data resources, its research value and commercial value is huge. In this paper, we use network node user’s influence to analyze this social network, and achieve very good results. This study has important theoretical and practical significance to micro-blog public opinion analysis.