Clustering analysis of e-commerce site users interest based on dissimilarity measure

The electronic commerce website's access log includes user's access data, through denoising analysis of the access data, the remained data contained the access user, the access time and the access link is carried on two-dimensional matrixing processing, the two-dimensional matrixing is calculated according to dissimilarity measure principle, the same access interest user are differentiated, the cluster of the electronic commerce website users is realized, which provides the reference for the website optimization and the commodity network promotion. Dur experiments analysis shows that the method is efficient and practical for the clustering analysis of electronic commerce website users' interest.