The Effects of Centrality Ordering in Label Propagation for Community Detection

Abstract

In many cases randomness in community detection algorithms has been avoided due to issues with stability. Indeed replacing random ordering with centrality rankings has improved the performance of some techniques such as Label Propagation Algorithms. This study evaluates the effects of such orderings on the Speaker-listener Label Propagation Algorithm or SLPA, a modification of LPA which has already been stabilized through alternate means. This study demonstrates that in cases where stability has been achieved without eliminating randomness, the result of removing random ordering is over fitting and bias. The results of testing seven various measures of centrality in conjunction with SLPA across five social network graphs indicate that while certain measures outperform random orderings on certain graphs, random orderings have the highest overall accuracy. This is particularly true when strict orderings are used in each run. These results indicate that the more evenly distributed solution space which results from complete random ordering is more valuable than the more targeted search that results from centrality orderings.

Dynamic community detection algorithm based on hyperbolic graph convolution

“…The experiments use the evaluation metrics NMI (Dickinson and Hu, 2015), ARI (Rand, 1971) and modularity Q (Newman, 2006) to evaluate the performance of the proposed algorithm. (NMI quantifies the similarity between two clusterings by assessing the mutual information shared between them.…”

Distributed Community Detection based on Apache Spark using Multi Label Propagation for Digital Social Networks

“…Community detection algorithms are classified into model-based, local optimized and label propagation algorithms. Numerous problems were reported related to time-efficiency in the first two categories of algorithms and the third is the most cited one i.e., label propagation methods such as LPA [9], SLPA [10,11], BMLPA [12] MLPA [13], DMLPA [14], LabelRank [15], LabelRankT [16] provides properties like low computational complexity, ease of implementation and accuracy in both disjoint and overlapping community detection. The mutual characteristic feature in the label propagation family is the process of exchanging community labels during propagation process between graph nodes in order to send the updated messages to neighbor nodes.…”