| With the rapid development of e-commerce network in China, more and more customersget their online shopping through e-commerce network. It is necessary and elementary formost e-commerce site reveal how to detect and predict the customer’s interest, and enhancethe customer’s internet experience.The concept lattice is the core data structure in the formal concept analysis theory, and apowerful tool for data analysis. With the growing surge of network data, the application of theconcept lattice becomes much more important.Using concept lattice theory to research on e-commerce is a common tool for solvingproblem with the e-commerce. Based on the theory of concept lattice with constraintsattribute, this dissertation provides a approach to analyze customs behavior on e-commercenetwork.At first, for customs interest mining in e-commerce network, this dissertation presents anapproach based on the property of the equivalence relation analysis, gives a quick equivalenceconcept lattice construction algorithm through the attributes of the concept lattice ofequivalence constraints. The algorithm quickly establishes the concept lattice of equivalentinterest. Meanwhile, it provides an example to illustrate the effective and practical of thisalgorithm.Afterwards, it deals with Web log that we obtained. The following step is to establishuser interest model utilizing the above equivalence concept lattice construction algorithm, andpresents customs interest forecasting methods. Applying this model and forecast methods intothe current behavior of the customs analysis, it will find out the custom’s interest, and makepredictions.Recommending the results to the customs, it not only increase the custom’s Internetexperience, but also it has high reference for the site operator to optimize web resource. |
