Clustering Web Transactions Based on Time Durations and Relationships

An effective method clustering web transactions using attribute information, and the relationship information between web transactions is proposed in the, paper. At first, each web transaction is transformed into a fuzzy vector with, the same length. Each element of a vector is a fuzzy variable representing the, time duration on a web page. Then we bring in a way to transform the relationship, between any two web transactions into a special attribute. Furthermore, Euclidean_distance, is adopted to measure the dissimilarity between any two new formed web transactions. At last, a numerical example is given to illustrate the clustering process., The results of the example demonstrates the given clustering method is more, meaningful.