The Online Recommended Shopping System Based On A Divided-and-improved Apriori Algorithm

Aiming at the online personalized recommendation system applied on webshopping, this study mainly discusses how to effectively use data mining techniques tomine association rules of all commodities from a large number of shopping transactionrecords, then recommends the appropriate product information to customers to helpthem find the really needed and useful product information.On the base of the analysis of web shopping characters, which represents the fieldof B2C e-commerce, it makes a comparation among different recommend techniquesin details and analyzes the applicability of these recommended technologies based oncontent, user information, association rules and collaborative filtering to B2C webshopping website. Based on this, a new automated recommendation system is raised,which describes how to combine the mechanism with data mining techniques andapply it to the recommended procedure. To make the system return betterrecommended results, an improved algorithm based on Apriori was proposed, and itwas demonstrated on the usage of web shopping website. Finally, this thesis analyzesthe lack of the actual use of this personalized recommendation system and raises somesolutions for B2C web shopping website.