Association Rules Algorithm Based Customer Relationship Management Applications

In order to show whole views between enterprises and customers sufficiently, process customer relationship management based on commercial sector is made develop to operational customer relation management based on organization and management functions. However, as customer information of enterprise and complex relation between enterprises and customers are fast increasing, present customer relationship management systems have not satisfied development of enterprises. To automatically extract potential and valuable commercial patterns from this complex trading information, it has become an inevitable trend for customer relationship management to develop from operation to analysis.For the past few years, the field of customer relationship management is occurring in the revolutionary change, the application of financial mathematics model and the development of computer technology have established a very good base for processing complex customer relationship. It provides a new moment for improving the level of customer relationship management that data mining technology based on data warehouse occurred. Hence, in the change of management pattern, of which center is from product to customer, we sufficiently use information resource of enterprise and connect data mining technology with customer relationship management system to find valuable commercial information from large trading data of enterprise, which can make management decision of enterprise become scientific to improve the timeliness of customer relationship management system. These are very important realistic significance for enterprise to face the fierce market competition.In order to find these associations between type of customer and feature of trading action in customer relationship management system, basing on association rule method of data mining technology and the analysis application of customer relationship management system, this paper proposes an algorithm of constraint association rules mining based on reuse attribute position indexing, expressed as MBRAPI, which can mine simple Boolean association rules that might satisfy constraint need of user and might be any long. Firstly, the algorithm uses the technology of reuse attribute position indexing to create index value candidate interval to thoroughly delete these candidate frequent itemsets which don't include constraint need of user. It can avoid the shortcoming that there are redundant candidate frequent itemsets in present similar algorithms to efficiently reduce computing of algorithm. Secondly, the algorithm uses index value candidate interval to generate candidate frequent itemsets to break the traditional idea of generating candidate frequent itemsets based on set theory. The algorithm reduces computing of generating candidate frequent itemsets and the number of scanned transaction data when computing support. Finally, via endpoints variable of index value candidate interval, the algorithm improves the efficiency of mining algorithm, which adopts binary complementing to double generate index value of candidate frequent itemsets to avoid computing that these similar algorithms turns integer into binary vector. This paper uses transverse and longitudinal test to compare the algorithm with present algorithms via simulate experimental platform. These results indicate that the executing efficiency of algorithm is faster and more efficient than typical algorithms and similar algorithms.To test practical applicability of the presented algorithm, this paper designs a customer relationship management system based on house sales. The algorithm is applied to the system to analyze data, namely, MBRAPI is used to mine association rules of features between customers and houses, aiming to different types of customs, the real estate enterprises can choose house position and improve house layout according to these associations, and do their best to improve customer satisfactions to make enterprise gain maximum benefits. Via these intelligent mining results of simulate system, we might find features of customers and their trading. These results might also test that the algorithm is practical in customer relationship management system.