Research On Tianjin Digital TV Customer Classificationbased On Python System

The main business and existing value-added service have become more and more mature with the rapid development of radio and TV industry.Under the constant development of tri-networks integration,the competition for customer resources in the domestic market becomes increasingly fiercer,and so the customer relationship management(CRM)plays a more important role in the operation management of radio and TV industry.It can reflect the true requirements of customers that the customers are segmented by studying the customers' behaviors,and thus providing support for the marketing decision of enterprises.Based on the analysis for the existing problems of customer segmentation in CRM of Tianjin Radio and Television,this thesis proposed the segmentation method for the consuming behaviors of customers by some relevant studies.Thus,the customer segmentation of Tianjin Radio and Television was obtained.First,this thesis stated the role of CRM in the operation management of enterprises and its organizational form,and thus the existing problems in the CRM of Tianjin Radio and Television were analyzed.Then the knowledge flow management in the CRM and its current situation in the Tianjin Radio and Television were also analyzed.Finally,the customer segmentation model was proposed based on the study on consuming behaviors of customers.Second,the relationship between the data mining technology and CRM was studied,as well as the advantages and disadvantages of K-means algorithm and clustering evaluation indexes were analyzed.Then,there formed the Python data mining platform based on K-means algorithm.Third,this thesis analyzed the existing problems in the customer segmentation of geographic grid of Tianjin Radio and Television and proposed the customer segmentation method based on RFM model.Then an idea to determine K value is put forward,as follows.In the first step,the scope of K value was determined according to the empirical value method.In the second step,the clustering results in combination with DBI index,and then finally,the results of customer segmentation and customer value analysis were obtained.It is concluded that the customers should be segmented to six groups by the comparison between empirical value method and DBI index method.Then,the customer value segmentation in the loyalty and stability of customer groups is completed.It is found by studies that there are more low-values customers for Tianjin digital TV as a whole as well as the existing value-added services and the marketing force of period products are poor,or the attraction is not strong enough to make customers purchase.So some new means would be needed,such as new media assets,promotion,advertisement,and so on.Moreover,it can provide meaningful support that the customer data are reasonably used for deep excavation.