Study And Application Of CRM Data Mining Based On Clustering Algorithms

Data mining is a rising crossover subject, involves an integration of techniques from multiple disciplines such as statistics, artificial intelligence, database technology. Clustering analysis is an important technology in data mining, which groups the data into classes so that objects within a class have high similarity in comparison to one another, but are very dissimilar to objects in other classes.CRM is the iteration of the procedure that translates the customer information into the positive customer relation. Customer classification is the grid-based method in CRM and people can recognize the useful information by using data mining technology.In this thesis, we discussed the concept, function type, processing procedure and technology algorithms of Data Mining. K-means is a partition in algorithms. The traditional k-means algorithm has sensitivity to the initial clustering center. Therefore, it may sink into the local minimum. To solve this problem, on the basis of analyzing the clustering result which relies on the starting value, the paper study the way of the starting value selection, a new method is proposed to find the initial clustering center. Meanwhile the papers also propose an improvement measure in regarding the computation of the clustering center which is sensitive to the isolated points. Based on customer's data of a company, we used above algorithm to have clustering analysis. The experimental results illustrate the practicability and effectiveness of the proposed method.