A Comparison of Community Detection Algorithms on Artificial Networks

Orman, Günce Keziban; Labatut, Vincent

doi:10.1007/978-3-642-04747-3_20

Cited by 85 publications

(96 citation statements)

References 40 publications

(86 reference statements)

Supporting

Mentioning

Contrasting

Order By: Relevance

“…In this section, we first describe the LFR model we applied during the first step, which supposedly allows generating the most realistic networks in terms of overall properties (degree distribution, smallworldness, etc.) [6,10,11]. Then, we shortly describe the community detection algorithms we selected, and explain how they differ on the way they handle the concept of community.…”

Section: Methodsmentioning

confidence: 99%

“…It increased the realism level even more by generating networks with power-law distributed degree. Among these newer models, we selected the LFR model, which seems to be the more realistic and was previously used as a benchmark to compare community detection algorithms [6,10,19]. The LFR model was proposed by Lancichinetti et al [6] to randomly generate undirected and unweighted networks with mutually exclusive communities.…”

Section: Network Generationmentioning

confidence: 99%

“…Additionally, networks with a community structure are characterized by a power-law distributed community size [9]. Several generative models with increasing realism were successively designed [6] before finally meeting these constraints [6,10,11]. However, recent studies showed other properties besides the community size distribution can be used to characterize real-world community structures [4,12], such as community-wise density and average distance, hub dominance, embeddedness.…”

Section: Introductionmentioning

confidence: 99%

See 2 more Smart Citations

Qualitative Comparison of Community Detection Algorithms

Orman

Labatut

Cherifi

2011

Communications in Computer and Information Science

Self Cite

View full text Add to dashboard Cite

Abstract. Community detection is a very active field in complex networks analysis, consisting in identifying groups of nodes more densely interconnected relatively to the rest of the network. The existing algorithms are usually tested and compared on real-world and artificial networks, their performance being assessed through some partition similarity measure. However, artificial networks realism can be questioned, and the appropriateness of those measures is not obvious. In this study, we take advantage of recent advances concerning the characterization of community structures to tackle these questions. We first generate networks thanks to the most realistic model available to date. Their analysis reveals they display only some of the properties observed in real-world community structures. We then apply five community detection algorithms on these networks and find out the performance assessed quantitatively does not necessarily agree with a qualitative analysis of the identified communities. It therefore seems both approaches should be applied to perform a relevant comparison of the algorithms.

show abstract

Section: Methodsmentioning

confidence: 99%

Section: Network Generationmentioning

confidence: 99%

Section: Introductionmentioning

confidence: 99%

See 1 more Smart Citation

Qualitative Comparison of Community Detection Algorithms

Orman

Labatut

Cherifi

2011

Communications in Computer and Information Science

Self Cite

View full text Add to dashboard Cite

show abstract

“…There are different community detection algorithms presented in the paper [1]. In this paper, we use a heuristic algorithm called a Louvian method [2] based on modularity optimization.…”

Section: Introductionmentioning

confidence: 99%

Anonymized Social Networks Community Preservation

Vadisala¹,

Kumari²

2017

ijacsa

View full text Add to dashboard Cite

Abstract-Social Networks have been widely used in the society. Most of the people are connected to one another, communicated with each other and share the information in different forms. The information gathered from different social networking sites is growing tremendously in large volumes of various research, marketing and other purposes which is creating security and privacy concerns. The gathered information contains some sensitive and private information about an individual, such as the relationship of an individual or group information. So, to protect the data from unauthorized users the data should be anonymized before publishing. In this paper, we study how the k-degree and k-NMF anonymized methods preserve the existing communities of the original social networks. We use an existing heuristic algorithm called Louvian method to identify the communities in social networks. We conduct the experiments on real data sets and compare the performances of the two anonymized social networks for preservation of communities of the original social networks.

show abstract

“…Optimisation algorithms include several approaches: greedy routing (Newman, 2004), simulated annealing (Guimera et al, 2004), spectral optimisation (Newman, 2006a), game-theoretic (Narayanam and Narahari, 2010;Chen et al, 2010), compression-based (Rosvall and Bergstrom, 2007) and flow-based (Wu and Huberman, 2004). Reviews of the various methods present in the literature can be found in Danon et al (2005) and Orman and Labatut (2009).…”

Section: Introductionmentioning

confidence: 99%

Label propagation algorithm: a semi-synchronous approach

Cordasco

Gargano

2012

IJSNM

View full text Add to dashboard Cite

Abstract:A recently introduced novel community detection strategy is based on a label propagation (LP) algorithm which uses the diffusion of information in the network to identify communities. Studies of LP algorithms showed that the strategy is effective in finding a good community structure. Label propagation step can be performed in parallel on all nodes (synchronous model) or sequentially (asynchronous model); both models present some drawback, e.g., algorithm termination is not granted in the first case, performances can be worst in the second case. In this paper, we present a semi-synchronous version of LP algorithms which aims to combine the advantages of both synchronous and asynchronous models. We prove that our models always converge to a stable labelling. Moreover, we experimentally investigate the effectiveness of the proposed strategy comparing its performance with the asynchronous model both in terms of quality, efficiency and stability. Tests show that the proposed protocol does not harm the quality of the partitioning. Moreover, it is quite efficient; each propagation step is extremely parallelisable and it is more stable than the asynchronous model, thanks to the fact that only a small amount of randomisation is used by our proposal.

show abstract

A Comparison of Community Detection Algorithms on Artificial Networks

Cited by 85 publications

References 40 publications

Qualitative Comparison of Community Detection Algorithms

Qualitative Comparison of Community Detection Algorithms

Anonymized Social Networks Community Preservation

Label propagation algorithm: a semi-synchronous approach

Contact Info

Product

Resources

About