Research Of Associated Clustering Collaborative Filtering Algorithm For E-commerce Recommendation System

With the rapid development of Internet technology, the e-commerce is also devel-oped rapidly. Online shopping has become a new form of shopping. With the develop-ment of e-commerce, Types of goods online are increasingly numerous, the e-commerce system structure become more and more complex. Choosing satisfactory goods from so many goods became a notable problem. Buyers choose satisfactory goods become very difficult and it is more difficult for sellers that they want to hold clients and make themselves impressively in the consumer’s mind. How to recommend their prefer goods for different users and how to improve the credibility of the seller becomes one of the key issues in e-commerce.In this paper, it mainly studies of the algorithm of associated clustering collabora-tive filtering(ACCF) and the e-commerce recommendation system. Firstly, the short comings of the traditional collaborative filtering algorithm is analyzed, Combining the association rules data mining algorithm with the collaborative filtering algorithm, it proposes the algorithm of associated clustering collaborative filtering. In this algorithm, fist of all, divides the goods into different categories by using the association rules technology, and then, maps the users who evaluated the goods into these categories. Then analysis the parameters, forms the users-items matrix which has been clustered, and then uses the collaborative filtering algorithm to form the recommendation. In this paper, it realizes the ACCF algorithm according to the previous analysis and design and uses the experimental data to prove the effectiveness of the algorithm. At the same time, this paper studies the model of the e-commerce recommendation system, analyses their workflow, analyses the requirement, describes the business of the system, designs the structure and the database of the system by combining the needs of this system and re-ferring to the software development process and specifications. And the system is di-vided into two modules of foreground service module and the background management module. The problem description, model designing and the implementation of the mul-ti-scenario recommendation modules and the pattern library creating module is given in this paper which sees as the emphasis of this paper.On the base of ACCF algorithm, in this paper, the e-commerce recommendation system recommend goods to users on different operating scenarios by the technologies of the association rules and collaborative filtering recommendation. In this article, it not only studies the theory of the data mining technologies and the recommendation tech-nologies, but also helps solve the reality problem of increasing user credibility and in-creasing sales. Therefore, this research has important theoretical significance and prac-tical value.