The Study Of Customer Knowledge Collection Method Based On Web Text Mining

In traditional Customer Relation Management(CRM), the customer-service-modal was adopted to deal with and solve problems of customers' information and knowledge. Under the guidance of thought above, the Customer Knowledge Collection(CKC) framework is established, providing mainly after-sale services. With the development of knowledge economy, advantages of information-oriented competition is transforming to those of knowledge-creation-oriented competition, therefore, implementation of Knowledge Management(KM) has been an important strategy of enterprises. Traditional customer-service-modal is fading out in Knowledge Collection.The author, facing the requirement of CKC, taking WEB-based customer knowledge as object, basing on methods and techniques of WEB Text Mining, establishes methods of WEB-Text-Mining-based customer knowledge collection and puts it into use, which is worthy of research in practices.Firstly, the differences and relations among Customer Knowledge Management(CKM), CRM and KM are discussed. According to the analysis of CKC process and three different kinds of customer knowledge, the author points out the deficiency of existing knowledge collecting methods and techniques, and brings forward the new framework of CKC process.And further more, taking the CKC requirement as a starting point, the author also suggests that WEB Text is the most valuable knowledge, thus researches on methods of WEB Text Mining. In the process, problems of existing WEB-Text-confronted KC methods, the project of semi-auto KC are all proposed in this paper.Then based on the analysis, an extracting rule of DOM-based WEB-Text, the induction algorithms and XML-based knowledge expressions are proposed to apply to treatment of WEB texts, and this further leads to the realization of the dynamic disposal and management of distributive knowledge.In the end, the WEB-Text-Mining-based CKC methods are applied to the design of software for Guangzhou Honda CKC Prototype System. WEB service, and "push" techniques are adopted here to design the procedure of search and subscribe, which add more functions to the CKC system. This work speeds up the realization of their CKC process. Comparing with the condition before, great effect has taken in many CKC aspects, which include auto motive parts, customer prejudice and satisfaction.