Data Mining Applied In Associated CRM

Associated CRM is applying for the company communicate with its customers which not only providing services but also collecting information from them.As the fast speed development of mobile Internet,it is much easier to connect companies and customers in various convenient ways.Data mining technology applying for CRM mainly for analyzing well-formed data in Analytical CRM which transmitted from Operational CRM,but high-capacity of original conversation data is neglected.This paper selected Sorting and Clustering algorithms which are widely applying in commercial analysis to sorting sessions from associated CRM to finding out the shape and regulation of data sets,improving the judge of products or services of the company,analyzing trends and potential requirements.Firstly,this paper introduced the development and application of CRM system,features of associated CRM and the requirements of the applying of its session data.To applying algorithm to session data,this paper also introduced common Sorting and Clustering algorithms and their lately researches.the following chapters improves the algorithms of Decision Tree to adapt CRM session data sets as these data sets not only have highly relevant but divergent as well.Secondly,to make multi-decision value conversation session be able to be sorted into different cluster,this paper proposed the multi-decision value sample split method.To reduce the tree generated time,this paper proposed another discretization partition method for continuous variable.Meanwhile,by reweight the row we make the data smoother and easier to cluster.In clustering phase,this paper proposed a variable k which is a preference parameter for affecting the amount of cluster,once when it do affect,the extra cluster will associate with density-based cluster as reference.Thirdly,the simulation result shows that comparing with original C4.5 algorithm,the new algorithm this paper proposed is both more effective,available and extensible which helps quality inspector to precise quality inspect and associated for business intelligence,recommended system and other data mining projects.