The Research Of User Needs Customization System Based On B/S Architecture And Web Mining

With the popularization and development of the Internet, electroniccommerce has been growing into people’s lives, how to optimize thesystem structure, improve system efficiency, to provide more and betterpersonalized service, which became an urgent e-commerce systemsproblem to solve.With the B/S (Browser/Server) architecture, we use SSHArchitecture to build a Customer demand customized system which useprovide personalized services Easy Installer. SSHframework is now used as an open source we application developmentframework, It includes the Struts, Spring and Hibernate they representthe presentation layer, business logic layer and data persistence layer, withSSH framework you can greatly increase software encapsulation andflexibility so that software becomes maintainable and scalable.In addition, to optimize the user experience, we will be using Webmining techniques, collected in the database the user needs analysis, data cleaning, sorting and conversion, and use the apriori algorithm to findassociation rules between the project package, in order to achieveselected according to user habits provide personalized customer service.the apriori algorithm is the most representative of an algorithm in datamining association rules. However, the apriori algorithm has two majordrawbacks:1it requires repeated several times to scan the database todetermine the number of candidate sets of support;2it takes too longtime to generate the frequent2-item sets. In this paper, we proposed anew algorithm to improve the operational efficiency of the algorithm.Easy Installer could manage all the applications released on givenproduct. It allows the application release owner import the applicationinformation to database, thus the image build could use the system exporttheir desired application lists to target media. The end user will use themedia to install all the applications on their client machine. Easy Installeris separated into2parts: application Import management, and applicationexport management, the input page part includes project informationmanagement (operating system environment, language, etc.), productname, version, and add archive and the corresponding silent installationcommand; the output page management section includes some of theexport package, the search part of the application package and to providefor customer choice. Due to the use of improved Apriori algorithm inEasy Installer The system will make data mining on user behavior, this would come to some rules which would recommend theappropriate products to users, and improve the user experience.