Construction personalized recommendation system of E-commerce based on variable precision rough set

E-commerce personalized recommendation technology is the core issue of automated recommendation system. The recommendation system of e-commerce is using data mining techniques to analyze visitor behavior in e-commerce site access, access to customers to help generate interest in the product information visit the recommended results. Variable precision rough set model is the expansion of the basic rough set, which is the basic rough set theory based on the introduction of the β that the relative error classification rate. The paper puts forward construct the personalized recommendation system of e-commerce based on variable precision rough set. The experiments show that the method can improve recommendation effective.