Link Clustering with Extended Link Similarity and EQ Evaluation Division

Abstract

Link Clustering (LC) is a relatively new method for detecting overlapping communities in networks. The basic principle of LC is to derive a transform matrix whose elements are composed of the link similarity of neighbor links based on the Jaccard distance calculation; then it applies hierarchical clustering to the transform matrix and uses a measure of partition density on the resulting dendrogram to determine the cut level for best community detection. However, the original link clustering method does not consider the link similarity of non-neighbor links, and the partition density tends to divide the communities into many small communities. In this paper, an Extended Link Clustering method (ELC) for overlapping community detection is proposed. The improved method employs a new link similarity, Extended Link Similarity (ELS), to produce a denser transform matrix, and uses the maximum value of EQ (an extended measure of quality of modularity) as a means to optimally cut the dendrogram for better partitioning of the original network space. Since ELS uses more link information, the resulting transform matrix provides a superior basis for clustering and analysis. Further, using the EQ value to find the best level for the hierarchical clustering dendrogram division, we obtain communities that are more sensible and reasonable than the ones obtained by the partition density evaluation. Experimentation on five real-world networks and artificially-generated networks shows that the ELC method achieves higher EQ and In-group Proportion (IGP) values. Additionally, communities are more realistic than those generated by either of the original LC method or the classical CPM method.

Discovery of Multi-Objective Overlapping Communities within Social Networks using a Socially Inspired Metaheuristic Algorithm

“…An extended link clustering method is presented in [4]. The algorithm runs extended link similarity and uses EQ metric for finding the top quality for the hierarchical clustering dendrogram part.…”

“…After getting biased, new candidates are specified. Power of each group is computed according to these candidates using equation (4). Inter-group competition starts after intra-group competition.…”

“…After the biasing process, regular members and candidates are reassigned. Powers of these groups are computed for new regular members and candidates using equation (4). These values are given in Table 11.…”

Research on key network node identification based on depth - first search algorithm

“…c be a child of 1 c , 4 c be a child of 2 c , 8 c be a child of 4 c , 5 c be a child of 8 c , and for a child of 5 c , let j be the child of node i, as in the above search order, Because the search sequence according to the above discussion, it has a neighbor node logic is no subsequent child nodes, so that it does not have children, and then 3 c is a child of 1 c , is a child node, the same 7 c is a child of 3 c , 6 c and 7 c has no child nodes. If j is a child of node i ,then A defines the search number of the earliest node in the subtree of j that can be traced back to node i by non-parent-side edges (i and j sides).…”

“…However, it is found that complex networks are not scale-based networks, but are not stochastic networks [2,3] . Therefore, the previous experience in identifying and protecting critical nodes is not suitable for complex networks with robustness, fragility and scale-free characteristics [4,5] , There is an urgent need for new methods and techniques to carry out research. In this paper, the key nodes in wireless sensor networks are probed, so it is assumed that the graphs formed by all the nodes and each node and its neighbor nodes are connected in advance, so we need to give the concept of graph traversal [6,7] .…”