The Design And Realization Of Goods Online Recommendation System Of Sports Goods Store

For a molding electric business enterprise, usually have the scale of tens of millions ofactive users, will produce large amounts of transaction data every day. But because of thecomplicated operation process, the enterprise internal has at present, most of the electricitybusiness enterprise also just for huge amounts of data type using simple addition, subtraction,multiplication, and division, a lot of data have been divided, not scattered data storage,integration, deep exploitation, such as structural analysis data in each operation link haven’tplay the important role of improve efficiency and reduce cost. The arrival of the age of thedata, the data to break the limitations of time and space in business provides a new way ofthinking, to "data liberation of productive forces" provides a new way.Through the application of data mining in electronic commerce, can provide users withpersonalized recommendation, for electricity customers added to the satisfaction degree of thestore, at the same time also can adjust the scheme and a guideline for electricity to providegoods, to enhance the competitiveness of the electricity, thus has good practicability.This paper studies the technique of data mining and the application of Web data miningin electronic commerce, according to the characteristic of e-commerce user, mainly adoptsclustering algorithm and association rules algorithm to realize the goods onlinerecommendations. First of all the information collected in the user database, businessinformation database and content information database, perform data preprocessing steps,including data extraction, cleaning, transformation and integration of data. Then execute suchas user identification, session identification and path identification to classify the dataaccording to the clustering algorithm further and to form the formation pattern libraryclassification cluster pattern mining, then according to the association rules algorithm ofgoods online recommendations set. Need in the process of the pattern matching algorithms are used to get the user might be interested in the commodity information, and feedback theresults through the server in the form of an HTML page to the user in the client browserinterface.In this paper goods online recommendation system, recommendation module is one ofthe most important function, when accessing the user to browse the website and issuerecommend commodities after the request to the server, will recommend the module of theuser in the browsing behavior of mode analysis, and to identify the session. Exists if the useraccess past records, it is will be assigned to the session request of existing clustering cluster, ifnot on the server to create a new cluster. Recommendation module, analyzing thecharacteristic of the user requests and access distribution corresponding to the user, clusterand based on the characteristics of the cluster set up goods targeted recommendations set, thengenerate the user can browse the HTML page, after a recommendation module sessionsshould be returned to the user to back the request.This paper design the sporting goods store design of the user registration, commodity list,product evaluation, product grades and recommend several functional modules such as online.1) The user registration. The user can choose to register and enter personal information,fill in the content of the information by the more comprehensive, is more convenient for theuser to classify, at the same time can also be more targeted to buy recommendations are putforward. User’s E-mail information can be sent in time according to the user preferences in thesale of commodity information.2) The commodity list. Store manager entry of goods entry all categories and all thegoods and list of prices of the goods, picture, name, etc, through the background. The user canchoose according to the price, the number of purchase, etc.3) Evaluation. User after buying a goods, can undertake assessment for the goods, theevaluation content can help other users as reference of purchase, but also conducive tomanagers to master user feedback of the product, and the classification, the price of acommodity, the respect such as online recommendations to adjust.4) Goods. The user can click on the commodity grading the goods they buy points isgiven.5) Online recommendations. This article mainly USES the association rules algorithm makes recommendations for the user’s browsing behavior, to buy records, according to theclustering method to cluster the user into, and according to the features of clusters wherespecific recommendations are presented.