The Improvement On CLIQUE Algorithm And Its Application & Research In Electronic Business Enterprises

Nowadays, with the rapidly development of e-business, e-business enterprises displayed its advantage and vitality. Many things including on-line shopping, on-line business, on-line agency is changing our life. Along with more and more competitions between e-business enterprises, the analysis of customers' registration&trading information is very important when it is used for forecasting customer behavior, finding out important customers, making decision for the enterprises.As a branch of statistic, clustering analysis is a kind of technology in data mining, which divide data into meaningful or useful cluster without experience and knowledge. Clustering analysis is a powerful tool when enterprises process their data by selecting right clustering algorithm to analyze data, mine potential & valuable information.As a grid-based algorithm, CLIQUE can make both all-sided clustering and subspace clustering .CLIQUE is suitable for the clustering of e-business data because of its good ability for processing data, but because it grid the data into a predefined grid ,this adds the complexity of the computation. Comparing with the CLIQUE algorithm, an improved algorithm is proposed in this paper, A set of experiments on synthetic dataset demonstrate the effect and efficiency of this improved algorithm.When it comes to application, using the data collected from China Ad net, we divide it into different clusters, and then we find out the useful rules by association algorithm which can be used by decision-makers to draw the management&selling strategy.