B2c Book Sales Site Of Association Rules-based Personalized Recommendation System

With the development of information and internet technology, E-commerce as a 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 made it hard for consumers to find the products and services they wanted, especially for them of B2C E-commerce. Because B2C E-commerce can provide enormous products and separate customers from real products.Personalized Recommendation systems emerged as the times require. By communicating with the users or analyzing their trade data, it can predict their preference and recommend products and services which they interest in. Personalized Recommendation systems can provide individualized services, increase cross-sell, extra entrance and Improve loyalty. With the help of this system, E-commerce companies can take advantage over their rivals.This thesis particularly chooses book sales B2C E-commerce website as a respective study object. Firstly, we analyzed the characteristics of different personalized recommendation technology. On the base of this analysis, thesis designed a recommendation system that suits for book sales B2C E-commerce website. Secondly, as for traditional Association Rule-based Recommendation algorithm's shortage, thesis utilized data partition conception and'Apriori'gained an efficient Association Rule-based algorithm. Using this algorithm we can mine customers'trade data and provide them personalized recommendation. Finally, thesis summarized the limitations of Association Rule-based Recommendation and presents some suggestion that may solve these problems.