Research On Structure Optimization Of E-commerce Website Based On Multi-objective Model

As the trading platform of e-commerce, website is the foundation of enterprise e-commerce’sdeveloping; its reasonable structure has very important impact on e-commerce’s operations. Areasonable website structure can not only help users saving time when they searching somecommodities which they were interested in, but also can bring more benefits and reputations tooperators. Therefore, made a research on e-commerce websites’ evaluation and optimization hasvery important realistic significance.This paper firstly introduced the related knowledge of e-commerce website and sometechnologies and methods involved in data mining, then evaluated the e-commerce website withthe evaluation method which is based on two-tuple linguistic. Webpage yield, webpageaccessibility and webpage loading time are used for the second time evaluation on the premise thatthe whole performance was reasonable. According to the website evaluation results, the websitestructure optimization problem is divided into two categories: one class is the optimization of pagestructure; another is the optimization of website link structure. In order to solving the problem ofpage structure optimization, this paper used association rules based on the external environment-combined sales optimization method and combination of collaborative filtering-personalizedrecommendation method to update the recommended list of combined merchandise sales andpersonalized commodity recommendation according to different website users. In order to solvethe link structure optimization problem, a multi-objective optimization model was proposed underthe premise of maintaining the basic website links, the model can be solved with tabu searchalgorithm. At last, the simulation experiment shows that the multi-objective optimization model isreasonable.Based on the related research of the e-commerce website evaluation and optimization, a multi-objective optimization model is established, further more association rules based on the externalenvironment-combined sales optimization method and combination of collaborative filtering-personalized recommendation method are used to update the recommended list of combined merchandise sales and personalized commodity recommendation, fulfilling the page structureoptimization.