Mining And Recommending Core Groups For Web Viral Marketing

Recently, The Web viral marketing is rapidly becoming an important marketing strategyin the field of E-business and a research focus of domestic and foreign experts. Because thecore groups play the important role in Web viral marketing, it has huge value of the research.However, the research of mining core groups is the beginning and there is lack of research inthe aspect of recommending for core groups domestic and overseas.First, a lot of information is looked up to understand the theory of Web viral marketing.We master the Web viral marketing and study the methods of mining core groups in-depthcurrently through results of previous studies.Secondly, we present the principles and methods to build Network-Influence Set of nodeon the basis of Web Customer Trust Network, and then present an algorithm called BUNIS(Building Up Network-Influence Set),and present an algorithm of mining core groups calledMCGNIS (Mining Core Groups Based On Network-Influence Set).Thirdly, a recommendation model called RMNIS (Recommending Model Based OnNetwork-Influence Set) and the corresponding algorithm are proposed in this paper. The stateand the mode of action in the Web viral marketing adequately are considered in RMNIS, itcombines the characters of individuation and no-individuation and present the formula tocalculate the recommendation degree.Finally, the experimental results show the Network-Influence Set can describe theintricate relationship among nodes of Web customer Trust Network well, and the algorithmsof mining core group based on Network-Influence Set using different thresholds has theadvantage of time or NC(Network-Coverage), RMNIS is accurately and diversely andsatisfies the business marketing purposes.