The advancement of computer、communication and network technology to promote therapid development of the Internet and e-commerce. At the same time the popularity of theInternet and the rapid development of e-commerce makes the amount of information overloadproblem highlights out, in the face of the vast amounts of product information users isdifficult to find to the products they need. E-commerce recommendation system is capable ofintelligent help users locate their goods quickly and accurately. However, this technology isstill has a series of problems such as data sparsity、system scalability and low accuracy.This article by means of sub-direction of the WEB data mining, together withe-commerce, proposed a model of e-commerce personalized recommendation system basedon WEB Data Mining. WEB data mining from WEB documents and WEB activities todiscover and extract interesting、potentially、useful patterns and hidden information, to meetthe future trends of the e-commerce.Innovation of this work is to combine the two techniques recommendations and parity,designed one online book parity recommendation system with the popular MVC developmentmodel. Different WEB data mining algorithms are researched and compared in the process ofimplement personalized browsing, finally, the main algorithm of the recommendation systemis selected is based on user and project double clustering collaborative filtering algorithm. Onthis basis, the two sub-modules, search for parity and personalized recommendation are toimplement. |