Study And Application Of Association Rules Mining In E-Commerce

With the development of information and internet techonology, E-commerce as an efficient new commercial model became more and more popular. Now people can complete business by tightly click the mouse. But E-commerce brings people not only convenience but also information overload. This situation makes it hard for consumers to find the products and services they wanted, especially for the consumers of B2C E-commerce. The techniques of data mining can solve this"Data explosion"problem well. Association rules mining, as an important branch of data mining, is simple and easy to explain and understand. It can describe the relationship between data efficiently. Mining association rules from large databases has become a hot area of research in recent years. You can find the instrinsic link between the commodity and commodity and also the commodity and customer through mining association rules. It has a very important guiding significance for the personalized recommendation in E-commerce, Enterprise market positioning and so on.This dissertation provides a detailed description of the association rules in the basic theory and algorithms of the classical algorithm- Apriori algorithm. Because in the field of E-commerce, the Apriori algorithm has some problems such as lower efficiency and redundant rules, this dissertation presents association rules mining algorithm based on rough set, as the characteristics of rough set theory, the algorithm can deal with the above issues better.This dissertation also focuses on the important trend of electronic commerce: personalized service, and introduced some popular recommendation technologies such as content based recommendation, collaborative filtering based recommendation and association rules based recommendation. In the dissertation it analyzes the Inadequate of these traditional recommendation technologies and presents a new recommendation technology, which bases on the ART Network and association rules. This recommendation technology not only can recommend the commodities by the associated items but also can recommend by the characteristic of the online user. At last of the paper, the recommendation mechanism is employed to implement a prototype E-commerce recommendation system to prove the feasibility and applicability of the presented method.