WEB-Mining-Based Recommender Systems In E-commerce

Web has become an important way to access information, since the growing Web information, people have to spend a lot of time to search, browse the information they need.Search engine (search engine) still can not satisfy the needs of different backgrounds, different purposes and in different periods for the request. Personalized Web services technology is made on this issue, which provided for different users of different services to meet different needs.Personalized Web is such kind of technology refers to the site of continuous learning experience, and improves the site's organization and the provision of information, in order to better provide information to users.In order to achieve personalized service, first of all, we needs to track and learn users' interested and behavior, and design an appropriate means of expression, and then recommend the resources users interested in to the user online. In order to recommend them more effectively and accurately, we must organize the resources, the characteristics of selected resources, and recommended the use of appropriate means.In this paper, from the perspective of application, we simply introduce a personalization technology development process of the Web-based Mining Web, presenting a number of excellent Web-based Mining Web personalization system. Then, this paper describes a Web-based personalized Web Mining processing, and from the direction of cluster analysis, association rules and sequence analysis model, it compared personalized Web Mining in the field of Web application technology status, from Web content Mining technology integration and the multi-features use to describe the characteristics of the corresponding technology trends.Based on the research Web personalization technology, this paper presented a framework of PRMD (Personalization Recommendation base Microsoft DMX) based on the analysis of the personality of DMX Recommended system. PRMD is such a recommendation system that integrated information use, content, and the subjective interests of the three aspects. It contains four processes: data acquisition, data pre-processing, data analysis and online recommendation.Then, in this paper, combining theory research and software application, to the practice of multi-angle mode, based in customer data and sale data of a supermarket chain, research in Customer Classify of supermarket, and purposed on custom behavior analysis of Customer. Finding problem in practice, and conclude experiences, and verify the model.In the paper, introduce the research content, development status and future trends of Data,data mining starting with the core content of recommended system. Build the model, OLAP and Clustering of the data using data mining. Aim at analysis the function of recommend system, and meantime providing a e-commerce decision analysis solution, to support the decision makers rapidly, accurately and comprehensive. At the same time, increasing customer value using information analysis, sharing information by applying advanced recommended system platform. Recommend system solutions can organize data and facilitate analysis to produce relevant information rapidly. And then extract meaningful rules from the information, to improve the quality of decision-making and shorten the time of solving the problem. At last, prospect the development of data mining in this paper. As the development of research and application, and demand-driven, data mining product and problem-sovling programme will be more intelligent, and also, the decision-making will be more easily.