Building Small World Networks by Clustering Analysis
One of the most important issues for Complex networks in nature is that how and why they have high clustering property, small world effect, and community structure. One possible principle is that complex networks are formed by similar behavior of their nodes. So Kmeans clustering, which is such a method to cluster objects by some measure of similarity, is used to construct the networks. The resulting networks are woven by all clustering paths and every clustering path is generated by following its clusters centroid. The experiments results of the resulting networks show properties of small world networks including small average distance and the clustering effect. Moreover the constructing process has also shown the evolution behaviors.
Jianyu Li Jie Zhou Rui Lv Xianglin Huang
Department of Automation Tsinghua University Beijing, China 100084 School of Computer Science Comm. Univ. of China Beijing, China 100024
国际会议
青岛
英文
2006-07-21（万方平台首次上网日期，不代表论文的发表时间）