The Research Of Online Shopping Mall Personalized Recommendation Modelbased On Data Mining

With the development of information technology and the popularity of the Internet, E-commerce has been characterized by rapid development in an efficient, convenient, low-cost manner. E-commerce website is becoming increasingly fierce competition, in the user-oriented today, when the user purchases goods, whether e-commerce website could meet their needs as soon as possible, have been the key to the success in the fierce competition. Based on this, the online store personalized recommendations become increasingly important.This paper study data mining, clustering, decision trees, and association rules algorithm recommended model in the online store application. First introduces the basic theory and related data mining algorithms for three algorithms used in this paper for a detailed description. From reality, this paper based on the data sets which to sell books online store to explore the data mining process. After finishing data processing, using tools such as SPSS Clementine, combined with clustering, association, decision tree algorithms to build four personalized model.First, established buying behavior prediction model in order to find key users according to the habits and keyword search access conditions. Then according to the user to access web content site, clustering and finding the most commonly purchased goods, build product recommendation model with C5.0 algorithm to evaluate the model. Next, according to the time interval of user clustering user first accesses the three pages and access between pages, and to identify possible access to the first four pages corresponding page recommendation. Finally, between the user’s basic information characteristics associated with commodity-style model, to identify characteristics associated with the user’s style and purchase of goods between the user of new visitors has similar characteristics to recommend appropriate style goods. For four model are evaluated and published results so as to further demonstrate the recommendation. Then it describes the application of personalized recommendation model, specifically describes how to embed the recommended model of e-commerce systems, and it may implement issues when recommendation function and the application process, finally pointed out the shortcomings of the paper and make some improvements.Sum up, through data mining project combines theoretical and practical, online mall to realize the construction of personalized recommendation model to solve problems online store personalized recommendations aspects of great significance. For data mining and recommendation system, the current study focused on improving the recommendation algorithm, the model will recommend how to form a personalized recommendation engine embedded in the study of small e-commerce site, the paper recommended model is embedded into the website system to achieve the function, accordingly introduced for future research in this area provides a guideline.