Clustering Algorithm In Economic And Trade Cooperation Between China And Russia

The daily log data of WEB site records lots of visiting information for the web. We can draw the user hobby information from log file. WEB site's designer proceeded the page reorganization and even intelligent web can predict the next visited page in the future. Clustering is an important area of application for a variety of fields including data mining and is an important method of data partition or grouping.Clustering is an important area of application for a variety of fields including data mining and is an important method of data partition or grouping. So far, there are 5 kinds of clustering algorithms including partition algorithm, hierarchical algorithm, density-based algorithm, grid-based algorithm and model-based algorithm.The paper first introduces classification of WEB mining techniques, with the emphasis on the algorithm used in this paper.K-means is a partition-based clustering algorithm, which divide n objects into K different kinds, of which, K is an input parameter. This algorithm clusters through continuous iteration, by which when the algorithm converges into an ending condition, iteration stops with the output of one cluster.The hierarchical method is the decomposition of the given data aggregate hierarchically. This method can be further divided into the agglomerative and schismatical cluster.Fuzzy cluster is actually the formation of a fuzzy matrix according to the properties of the research subject, and the confirmation of the categorization relation in view of the degree of membership. This paper gives a detailed description of the fuzzy algorithm used in science literature. For new users, when they browse the website for the first time, classification can be made by computing the similarity between the new and other users.Lastly, the writer conducts analysis of the application of K-Means and hierarchical cluster in the diary analysis system of the Sino-foreign Economic and Trade Web, clustering diary data of the web and doing theoretical transplantation of the fuzzy cluster algorithm in preparation for the future personalized services.