Finding Tenuous Groups in Social Networks

Abstract

As real networks become larger and larger, finding groups in a network has become necessary in real applications. While current researches mostly focus on finding dense subgroups with dense relationships, finding tenuous groups (abbrev., TGs) has not got much attention. Tenuous groups, which are sets of nodes with few interactions or weak relationships, also have many practical applications and great significances. In this paper, we model the problem of finding tenuous groups to the problem of finding sets of nodes where the shortest distances between most pairs of nodes are greater than a given k. We first introduce a formal definition of K-Line-Minimized (abbrev., KLM) problem. Then, we propose an efficient K-Line-Minimized-Algorithm (abbrev., KLMA) for KLM Problem. We conduct extensive experiments and comparisons to demonstrate that the proposed algorithm works well in solving TGs problem and outperforms another related algorithm for Minimum-K-Triangle-Disconnected-Group (abbrev., MKTG) problem in terms of efficiency and effectiveness.