The Study Of Text Mining And Application In CRM System

The implementation of customer relationship management(CRM) have an important role to improve the core competitiveness of enterprise, especially for the customer service center which provides the service to customer, CRM system can help companies improve work efficiency and enhance service competitive strength. With the customer service CRM system gradually application and increase the customer data. Use the information processing technology to handle massive intelligence information is the inevitable trend of the development of the CRM system.At present, data mining was more applied in the CRM system, mostly using data mining algorithms for classify customer data to segment customers, or use association rules mining to the products to realize cross-selling etc. Text mining is mainly being used on the classify internet web page,text retrieval,filter junk e-mail,etc, the application of text mining also less in CRM system which have more technical document. This paper realized the application of text mining in customer service CRM system which based on the study of existing CRM and text mining technology. Particularly, the research content of this paper included the following areas:Firstly, this paper describes the research situation of CRM and text mining, analyzes the benefits to enterprise after implementation the CRM, and briefly describes the definition of data mining, the classification of data mining and current research focus, thus leads to the text mining technology.Secondly, research of the key technology on text mining, and focuses on investigate the feature item weight. This paper analyzed the shortage of the traditional TFIDF algorithm, and then proposed a feature item weight algorithm based on keywords synonymy replace and adjacent merger. The KSRAM algorithm was used to extract the keywords experiments; the result shows that the KSRAM algorithm was better than TFIDF algorithm in precision and recall.Thirdly, this paper did all-sided analysis to service and support CRM system. The text mining technology was used in fast acquire solutions module, and the process of fast acquire solutions was regarded as the process of text retrieval. Then gived a specific example, the result shows that using the KSRAM algorithm to get more knowledge in precision than TFIDF algorithm.