Font Size: a A A

Research And Application Of Data Mining For Telecom Customer Relationship Management System

Posted on:2008-01-24Degree:MasterType:Thesis
Country:ChinaCandidate:W X TanFull Text:PDF
GTID:2178360215990057Subject:Computer technology
Abstract/Summary:PDF Full Text Request
At present, as the situation of monopolization has broken down, competition among the national telecommunication companies is more and more furious, whereas, the difference of network service is reducing from time to time. As a result, competing rivals will have to surfer the loss due to merely the single price war. Almost all the telecommunication companies are looking for a good solution to improve their service quality as well as the competitive ability. These telecommunication companies have to improve the internal scientific decision-making ability and meanwhile strengthen the judgment ability urgently to face more severe competition in the market. That's why an effective Customer Relationship Management (CRM) System is in great demand.Actually, the telecommunication companies have accumulated a mass of digitalized operation data. They can search out a lot of valuable information from user data with the aid of data extraction techniques, take custom consumption characteristics for example. Such data can be further interpreted by the marketing department to provide more powerful marketing strategy and also reduce a certain cost. Therefore, a mass of electronic data has laid a technology foundation to build up the Customer Relationship Management System for telecommunication companies.In order to improve service quality, survive market competitions and excavate operation data, China Telecom Jiangmen Company is trying to build up the customer relation management system and to widen business application possibilities from time to time by investigating and using their data extraction technique. The research in this dissertation orientates on the background of"Jiangmen Telecom CRM Assistant System"developed by China Telecom Guangzhou Research and development Center, China Telecom Guangdong Ltd. Technology Academe and China Telecom Jiangmen Company aim to set up a telecommunication customer loss forecast model.This paper makes the primary investigation to automatically breakdown customer group by a clustering arithmetic. Meanwhile, according to history data such as customer background information, consumption/payment details, behavior data and relative instructional criteria etc., the author ties to explore customer loss forecast required by the CRM through the data extraction technique like employing the decision-making tree. As to the problem raised in the current relative investigation, a detailed solution, including the customer description, attribute stipulation, loss reason analysis & forecast and control strategy etc., is provided to relieve major concern on customer loss. Attention is drawn to (1) solve construction problem of the loss model ; (2) apply Logistic regressive arithmetic & nerve network technology to supervise the study model parameter; (3) improve model training and precision & reliability of the forecast; and (4) use the post-training model to forecast potential customer loss so as to create a possible losing customer list for the company.Based on case study in real applications, the author has further proved the validity and effectiveness of this model. Its significance lies on foreseeable practical instructions as well as potential marketing strategy for telecommunication companies.
Keywords/Search Tags:Data Mining, Customer Relationship Management, Customer Breakdown, Clustering, Decision Tree
PDF Full Text Request
Related items