The Application Of Data Mining On Value-Added Services

In the context of deepening the electronics, the competitive level of mobile communications market is also rising, at the same time, a relatively new science and technology-data mining has gradually come into our field of vision, it provides a great convenience for the era of big data. In recent years, the application scope of data mining is increasing quickly, and the research value is also growing, it has been sought after by many domestic and foreign researchers.In2012, the chinamobile report shows that the total income of the mobile value-added services accounted for main business income30%,compared with2011,there is the growth of4percentage points, it shows that the value-added services is already one of the main source of income, how to operators these value-added services plays a vital role for the future development of mobile. This paper is mainly to analyze and mine the using circumstances of the mobile value-added services users from the following two aspects.Firstly, through the application of Clementine, we start to perform linkage analysis to nine mobile business (namely mobilegames,m obilevideo, MM, mobilereading,139mailboxs, fetion, mobile news, the whole song and mobile surfing), the analysis is used to discover the relevance among the active business of chinamobile and then applying this business relevance to recommend business to the users. The whole analysis is to achieve two goals, finding recommended businesses under the attributed phone numbers, and according to the recommendation of the chosen business, to find the appropriate user for marketing.Secondly, in the whole users of mobile value-added services, there exist a large proportion silent users, here, with regard to this problem, we carry out a decision tree model(here we select the business fetion),in order to analyze what type of the users can easily active the business, according to the decision tree model, we can finally acquire a serious of decision rules, then according to the decision rules, we can easily pick out the users who may active the business to activate purposefully.