A Unified Semi-Supervised Community Detection Framework Using Latent Space Graph Regularization

Abstract

Community structure is one of the most important properties of complex networks and is a foundational concept in exploring and understanding networks. In real world, topology information alone is often inadequate to accurately find community structure due to its sparsity and noises. However, potential useful prior information can be obtained from domain knowledge in many applications. Thus, how to improve the community detection performance by combining network topology with prior information becomes an interesting and challenging problem. Previous efforts on utilizing such priors are either dedicated or insufficient. In this paper, we firstly present a unified interpretation to a group of existing community detection methods. And then based on this interpretation, we propose a unified semi-supervised framework to integrate network topology with prior information for community detection. If the prior information indicates that some nodes belong to the same community, we encode it by adding a graph regularization term to penalize the latent space dissimilarity of these nodes. This framework can be applied to many widely-used matrix-based community detection methods satisfying our interpretation, such as nonnegative matrix factorization, spectral clustering, and their variants. Extensive experiments on both synthetic and real networks show that the proposed framework significantly improves the accuracy of community detection, especially on networks with unclear structures.

Investigating community structure in perspective of ego network

“…Various cut based network level properties such as ratio cut (Leighton & Rao, 1988;Wei & Cheng, 1989), normalized cut (Shi & Malik, 1997) and conductance (Brandes et al, 2003) have been defined. Special class of community detection algorithms known as spectral clustering (Nascimento & de Carvalho, 2011;Yang, Cao, Jin, Wang, & Meng, 2015) uses mostly these properties to partition the network. Several other approaches such as Optimization Methods (Duch & Arenas, 2005;Handl & Meyer, 2007;Schaeffer, 2007) and Bayesian Methods (Sakya & Biswas, 2012) also use network level properties.…”

Improving the Efficiency and Effectiveness of Community Detection via Prior-Induced Equivalent Super-Network

“…Given in advance the pairwise prior information, passive semi-supervised community detection designs the algorithm to increase the performance as much as possible 10–14 . For example, Zhang et al .…”

“…Yang et al . unify many existing community detection algorithms, including nonnegative matrix factorization and modularity maximization model into a clustering framework in latent space 14 . To force a pair of nodes with must-link to belong to the same community, they encode them to have similar latent space representations by introducing a weighted latent space graph regularization.…”

“…Third, most methods require tuning additional parameters to ensure effective encoding of the pairwise prior information. For example, Yang’s unified semi-supervised framework introduces a parameter balancing the impact of the topology information and priors to maximize the performance improvement 14 . All of these drawbacks limit the application of these methods for problems.…”

Semi-supervised Community Detection Framework Based on Non-negative Factorization Using Individual Labels

“…These algorithms use the prior information by transferring and modifying the adjacency matrix directly. After reconstructing the adjacency matrix, the semi-supervised problem is transformed into an unsupervised one [10].…”