Research on Automatically Clustering Algorithm in Web Personalize Service
Zhang Wang · Microelectronics & Computer · 2007
There is the defect of needing user to input cluster number in the traditional K-Means clustering algorithm,this paper improves the clustering algorithm,proposed a validity index to overcome this problem. This algorithm realizes automatically cluster,which improves cluster validity without background knowledge. And this algorithm can also choose appropriate step based on the scale of the data,strengthen its adaptability. This paper uses the improved algorithm to cluster web users in the data of web log,and verifies this algorithm.