Graph Convolutional Networks Meet Markov Random Fields: Semi-Supervised Community Detection in Attribute Networks

Jin, Di; Liu, Ziyang; Li, Weihao; He, Dongxiao; Zhang, Weixiong

doi:10.1609/aaai.v33i01.3301152

Cited by 110 publications

(52 citation statements)

References 14 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…In the future, we will also attempt to extend LSSGA with deep learning (e.g., Graph Convolutional Networks(GCN) [54] and variational autoencoder [55]), to detect the community structure of attributed networks whose nodes usually have one or more attributes, which are equally important to the topological structure in community detection.…”

Section: Discussionmentioning

confidence: 99%

“…To realize community detection, community detection methods, such as heuristic- [1], [10]- [22] and optimization-based algorithms [23]- [43] have been proposed. Moreover, advanced deep learning-based algorithms [54], [55] have been proposed in recent years. Among them, evolutionary computation-based algorithms are the most widely used techniques in community detection.…”

Section: Introductionmentioning

confidence: 99%

See 1 more Smart Citation

Community Detection Based on Genetic Algorithm Using Local Structural Similarity

Guo

Zhou

et al. 2019

IEEE Access

View full text Add to dashboard Cite

Community detection is an important research direction in complex network analysis that aims to detect the community structure in networks via clustering operation, which has an important application value and practical significance in mining potential network information. Genetic algorithm (GA) is commonly used in community detection to solve NP-hard problems caused by modularity optimization effectively. However, in terms of GA, the problems of global search performance and slow convergence still remain due to the low accuracy of initial population generation and the randomness of the mutation operator. Thus, a novel local structural similarity-based GA (LSSGA) is proposed in this study to address such problems. It mainly contains two innovations: First, a new generation strategy for initial population based on local structural similarity and roulette wheel selection is designed to improve the quality of initial individuals, at the same time, maintain the diversity of the initial population. Secondly, an effective mutation operator based on label propagation and local structural similarity is proposed, which can quickly achieve targeted and effective mutation. The performance of LSSGA has been verified in many detailed aspects such as the convergence of initialization strategy, the effectiveness of overcome falling into local optimum, and rationality of objective function. Experimental results of both synthetic and real-world networks demonstrate the effective performance of the proposed method via systematic comparison with other existing advanced algorithms.

show abstract

Section: Discussionmentioning

confidence: 99%

Section: Introductionmentioning

confidence: 99%

Community Detection Based on Genetic Algorithm Using Local Structural Similarity

Guo

Zhou

et al. 2019

IEEE Access

View full text Add to dashboard Cite

show abstract

“…Then used that as an a priori probability, compute the posterior probabilities by a loopy belief propagation algorithm over the MRF, to, finally, optimizing the belief propagation algorithm by a second neighbor criteria that sparsifies the adjacency matrix. Further optimization of similar ideas was obtained by using graph convolutional networks, i.e., CNNs over CRFs [177]. Attribute inference in social network data via MRFs can also be used to improve cybersecurity algorithms [178], to learn consumer intentions [179], to study the epidemiology of depression [180] among other issues.…”

Section: Applications Of Mrfs In the Analysis Of Social Networkmentioning

confidence: 99%

Random Fields in Physics, Biology and Data Science

Hernández‐Lemus

2021

Front. Phys.

View full text Add to dashboard Cite

A random field is the representation of the joint probability distribution for a set of random variables. Markov fields, in particular, have a long standing tradition as the theoretical foundation of many applications in statistical physics and probability. For strictly positive probability densities, a Markov random field is also a Gibbs field, i.e., a random field supplemented with a measure that implies the existence of a regular conditional distribution. Markov random fields have been used in statistical physics, dating back as far as the Ehrenfests. However, their measure theoretical foundations were developed much later by Dobruschin, Lanford and Ruelle, as well as by Hammersley and Clifford. Aside from its enormous theoretical relevance, due to its generality and simplicity, Markov random fields have been used in a broad range of applications in equilibrium and non-equilibrium statistical physics, in non-linear dynamics and ergodic theory. Also in computational molecular biology, ecology, structural biology, computer vision, control theory, complex networks and data science, to name but a few. Often these applications have been inspired by the original statistical physics approaches. Here, we will briefly present a modern introduction to the theory of random fields, later we will explore and discuss some of the recent applications of random fields in physics, biology and data science. Our aim is to highlight the relevance of this powerful theoretical aspect of statistical physics and its relation to the broad success of its many interdisciplinary applications.

show abstract

“…Compared with traditional community detection approaches, deep learning-based methods aim to identify community structures by creating more powerful representations of node attributes and community structures [27]. Concretely, depending on the used learning strategies, deep learning-based methods for finite and infinite community detection fall into five main categories: convolutional neural network (CNN)based [28], auto-encoder-based [29], generative adversarial network-based [30], graph embedding-based [31] [32] and graph neural network (GNN)-based [33] [34]. Comprehensive surveys [27] [35] of community detection approaches are referred.…”

Section: A Key Structures Detectionmentioning

confidence: 99%

Learning Concept Interestingness for Identifying Key Structures From Social Networks

Gao

Hao

Pei

et al. 2021

IEEE Trans. Netw. Sci. Eng.

View full text Add to dashboard Cite

Identifying key structures from social networks that aims to discover hidden patterns and extract valuable information is an essential task in the network analysis realm. These different structure detection tasks can be integrated naturally owing to the topological nature of key structures. However, identifying key network structures in most studies has been performed independently, leading to huge computational overheads. To address this challenge, this paper proposes a novel approach for handling key structures identification tasks simultaneously under the unified Formal Concept Analysis (FCA) framework. Specifically, we first implement the FCA-based representation of a social network and then generate the fine-grained knowledge representation, namely concept. Then, an efficient concept interestingness calculation algorithm suitable for social network scenarios is proposed. Next, we then leverage concept interestingness to quantify the hidden relations between concepts and network structures. Finally, an efficient algorithm for jointly key structures detection is developed based on constructed mapping relations. Extensive experiments conducted on real-world networks demonstrate that the efficiency and effectiveness of our proposed approach.

show abstract

Graph Convolutional Networks Meet Markov Random Fields: Semi-Supervised Community Detection in Attribute Networks

Cited by 110 publications

References 14 publications

Community Detection Based on Genetic Algorithm Using Local Structural Similarity

Community Detection Based on Genetic Algorithm Using Local Structural Similarity

Random Fields in Physics, Biology and Data Science

Learning Concept Interestingness for Identifying Key Structures From Social Networks

Contact Info

Product

Resources

About