An Application Of Improved Hits Algorithm In User Influence Valuation Sstem Of SNS Websites

Internet technology brings human beings into the information age, This changes alsoeffect in people’s social network. Especially the birth of Web2.0products recent years,Facebook as representatives of a number of social networking products,brings a variety ofelements in the reality of human relationships. The influence and network resources ofdifferent users directly shows the tremendous commercial value and spread of potentialitself. How to measure the actual influence of different users effectively, and then tap thepotential value in it has become an urgent problem.In this paper, we proposed a method to evaluate user influence of SNS (SocialNetworking Services) websites based on the HITS (Hyperlink-Induced Topic Search)algorithm. HITS algorithm is a measure to evaluate pages authoritative in Web data miningalgorithm,we introduced the algorithm to the evaluation of the influence of SNS users, andimproved it based on the characteristic of SNS site itself, in order to enable it to adapt theneeds of evaluate the user influence. First, we t introduce the measure of adding correctionvalue of Hub according to the style of users In order to avoid the error concerned about thenumber of weight problem; Second, against the dispoint of HITS algorithm which does notconsider the contents of the authority, we improved a method of content authority coefficient,which add the contents into the influence evaluation.Finally, we introduced the concept ofnetwork blocking degree based on other factors of the SNS site, so that can further amendedthe result.In this article we first introduced the research background and significance, as well as thespecific work of this article and article structure; Second, we introduces the basic concept ofsix degrees of separation theory, and the SNS, the theory, development status and the currentmainstream form of SNS service, and their respective characteristics. SNS product networkhas a potential value of data mining, which as the basis for the improvement work; once againwe proposed the HITS algorithm improvement program, and implied the evaluation of theuser influence the algorithm. In this section,we considered the influence of SNS sites socialnetworking as three parts, namely: the user structure authority, the user contents authority and network blocking degree. The user structure authority is to quantify the improved focus,Inthis article, we propose a different weight correction methods based on different types of usersto the Hub values given, in addition to user network blocking coefficient and the contentauthority coefficient, and finally get ultimately influence evaluation; then we designed asimple user influence evaluation system according to the above improved algorithm withJava. Among which the system random sampling from Sina microblog users group, to collectthe relevant data of its users to evaluate the influence of users and sort,based on the improvedHITS algorithm; In addition, In order to verify the influence evaluation system is effective,wealso introduced the result from sina evaluation system.compared which.with the result fromthe user questionnaire Finally we carried out summary and outlook.