Research And Design Of Customer Relationship Management System Based On Data Mining Techniques

Data mining is the process of extract information and knowledge from thedata, which are numerous, incomplete, noise, vague and random. The extractedinformation is hidden, unknown, and potentially useful.Mainly in the following five categories of data mining functions:classification, clustering, association rules and sequential pattern discovery,prediction, deviation detection. The main aspects of data mining applicationsinclude: financial, securities, insurance, and telecommunication, retail trade,biomedical and DNA analysis, electronic commerce.Classification is the most common data mining tasks. Classification is tounderstand the characteristics of things and make predictions using historical datato build a classification model process.The main purpose of classification is to analyze the input data, through thetraining focused on the characteristics of the data shown for each class to find anaccurate description or model. Classification of the general steps is creating themodel, then model uses.The quality assessment of classification based primarily on predictionaccuracy, computational complexity, the model described in simple degrees.Commonly used classification of data methods: decision tree induction,Bayesian networks, neural networks, genetic algorithms, rough sets and so on.There are many decision tree algorithms; this paper describes the ID3, SLIQand the SPRINT algorithm.Customer Relationship Management (CRM) is to help enterprises to maintain their competitiveness, retain customers and maximize profits from their customers,reduce marketing costs. Its main content is the detailed information to customersthrough in-depth analysis to increase customer satisfaction, thereby enhancing thecompetitiveness of enterprises.Of customer data collection, analyze the relationship that exists hidden in thedata and rules, in the vast database to find useful information, often need to usedata mining techniques. Data Mining in Customer Relationship Management canoffer businesses a full range of management and better customer communicationskills, maximize customer profitability.In the access to relevant information, based on meticulous research and datamining technology, data mining algorithms in the decision tree algorithms,customer relationship management (CRM) and data mining in CRM applicationand other content, clear data in customer relationship management the applicationmainly in the following areas:(1) Customer value analysis. Use the classification or clustering method todivide the value of the customers.(2) Product analysis of customer value. Basic methods and customer valueanalysis is the same.(3) Customer retention. With the clustering (classification) and onlineanalytical processing (OLAP), customer base can be divided into five categories:stability, high-value customer base, the loss of high-value customer base easy,low-value and stable customer base, the loss of low-value easy to customer base,there is no value customer base.(4) Customer credit analysis.(5) The customer satisfaction analysis. Primarily to help companies improvetheir customer marketing strategies, resulting in increased customer satisfaction,increase customer loyalty.In the end, use the data mining techniques in customer relationshipmanagement, customer relationship development of the advertising management system. The system adopts the B/S structure, based on Ruby on rails frameworkand Web2.0 development. Data mining is used in decision tree algorithms, onlineanalytical processing (OLAP) and data warehousing.The customer relationship management system is divided into six modules:system management, customer management, availability management, andrecommended Listing, report management, and call center. Recommended in theListing are used in matching algorithms. The greatest feature of this system is thatcan be more objective, intelligent formation of availability of information, saving alot of human and material resources to enhance the real estate agents thecompany's operating efficiency, customer will have to bring the greatest. In a laterstudy, also the improvement of intelligent recommendation algorithm, in order toimprove efficiency of operation and recommended for accuracy.