The Research Of Clustering Analysis In Telecom Customer Segmentation

In telecommunications, because of customer data increasing explosively, cluster analysis of data mining technology has been taken into the telecommunications customer relationship management in order to advance the decision-making efficiency. In data mining, most objects have no strict class attribute and subordinate relationship, they have the overlaps, cross in attribute aspects, so it is more suitable for fuzzy division, and because many objects are fuzzy with each other, so the fuzzy clustering analysis is produced. Fuzzy clustering is applied in every sector has become a hotspot.The contents of this article include the following aspects:Firstly, this article includes the related knowledge and technologies about data mining, clustering analysis, fuzzy theory, and it is significance that fuzzy is applied to data mining, and analyzes and describes the clustering algorithm by using transitive closure based on fuzzy similarity relation to calculate fuzzy equivalence relations, and on this basis, proposed a new fuzzy clustering algorithm which considers the weighting factors, so that the assessment and classification results will be more ideal. For the determination of weighting factors, we integrated the use of information entropy and the coefficient of variation method. Then, the analysis of the clustering based on subtractive clustering and cluster validity evaluation, analyzes and verifies the clustering validity. and finally, introduces the knowledge of customer segmentation, and the research content is applied to the customer segmentation, and processes the telecommunications simulated computation, the simulation results show the feasibility and effectiveness of the algorithms.