A review of clique-based overlapping community detection algorithms

Gupta, Sumit; Singh, Dhirendra Pratap; Choudhary, Jaytrilok

doi:10.1007/s10115-022-01704-6

Cited by 11 publications

(7 citation statements)

References 87 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…More precisely, the proposed approach aims to discover densely relevant overlapping multi-communities, since it is based simultaneously on structural information and the similarity of the relevant dimensions. In fact, it has been proven by Gupta in [18] that, the overlapping form of structures containing triads (being one of the considered structural information) helps in the detection of optimized covers. Definition 2.…”

Section: Relevant Covermentioning

confidence: 99%

UCAD: commUnity disCovery method in Attribute-based multicoloreD networks

Domgue,

Tsopze,

Ndoundam

2024

Preprint

View full text Add to dashboard Cite

Many hierarchical methods for community detection in multicolored networks are capable of finding clusters when there are interslice correlation between layers. However, in general, they aggregate all the links in different layer treating them as being equivalent. Therefore, such aggregation might ignore the information about the relevance of a dimension in which the node is involved. In this paper we fill this gap by proposing a hierarchical classification based-Louvain method for interslice-multicolored networks. In particular we define a new node centrality measure named \textit{Attractivity} to describe the inter-slice correlation that incorporates within and across-dimension topological features in order to identify the relevant dimension. Then, after merging dimensions through a frequential aggregation, we group nodes by their relational and attribute similarity, where attributes correspond to their relevant dimensions. We conduct an extensive experimentation using seven real-world multicolored networks, which also includes comparison with state-of-the-art methods. Results show the significance of our proposed method in discovering relevant communities over multiple dimensions and highlight its ability in producing optimal covers with higher values of the multidimensional version of the modularity function.

show abstract

Section: Relevant Covermentioning

confidence: 99%

UCAD: commUnity disCovery method in Attribute-based multicoloreD networks

Domgue,

Tsopze,

Ndoundam

2024

Preprint

View full text Add to dashboard Cite

show abstract

“…Communities can be made up of one or more building blocks, and a building block can be a self-contained community or not. The definition of a community is quite general, and it also allows for hierarchical and overlapping communities [10,[18][19][20][21][22]. Our approach, which involves searching for local maxima of the quality function, is well suited for identifying complex and overlapping community structures.…”

Section: Introductionmentioning

confidence: 99%

Detecting Overlapping Communities Based on Influence-Spreading Matrix and Local Maxima of a Quality Function

Kuikka

2024

Computation

View full text Add to dashboard Cite

Community detection is a widely studied topic in network structure analysis. We propose a community detection method based on the search for the local maxima of an objective function. This objective function reflects the quality of candidate communities in the network structure. The objective function can be constructed from a probability matrix that describes interactions in a network. Different models, such as network structure models and network flow models, can be used to build the probability matrix, and it acts as a link between network models and community detection models. In our influence-spreading model, the probability matrix is called an influence-spreading matrix, which describes the directed influence between all pairs of nodes in the network. By using the local maxima of an objective function, our method can standardise and help in comparing different definitions and approaches of community detection. Our proposed approach can detect overlapping and hierarchical communities and their building blocks within a network. To compare different structures in the network, we define a cohesion measure. The objective function can be expressed as a sum of these cohesion measures. We also discuss the probability of community formation to analyse a different aspect of group behaviour in a network. It is essential to recognise that this concept is separate from the notion of community cohesion, which emphasises the need for varying objective functions in different applications. Furthermore, we demonstrate that normalising objective functions by the size of detected communities can alter their rankings.

show abstract

“…The problem of community detection is NP-hard, and unlike graph partitioning, the number and size distribution of communities is not known beforehand. There are various methods for community detection, including label propagation [15,14,16], random walk [17], diffusion [18], spin dynamics [19], fitness metric optimization [20,21], statistical inference [22,23], simulated annealing [24,19], clique percolation [25,26,27], and more. These methods can be divided into two main groups: divisive and agglomerative.…”

Section: Introductionmentioning

confidence: 99%

Selecting a suitable Parallel Label-propagation based algorithm for Disjoint Community Detection

Sahu¹

2023

Preprint

View full text Add to dashboard Cite

Community detection is an essential task in network analysis as it helps identify groups and patterns within a network. High-speed community detection algorithms are necessary to analyze large-scale networks in a reasonable amount of time. Researchers have made significant contributions in the development of high-speed community detection algorithms, particularly in the area of label-propagation based disjoint community detection. These algorithms have been proven to be highly effective in analyzing large-scale networks in a reasonable amount of time. However, it is important to evaluate the performance and accuracy of these existing methods to determine which algorithm is best suited for a particular type of network and specific research problem. In this report, we investigate the RAK, COPRA, and SLPA, three label-propagation-based static community discovery techniques. We pay close attention to each algorithm's minute details as we implement both its single-threaded and multi-threaded OpenMP-based variants, making any necessary adjustments or optimizations and obtaining the right parameter values. The RAK algorithm is found to perform well with a tolerance of 0.05 and OpenMP-based strict RAK with 12 threads was 6.75x faster than the sequential non-strict RAK. The COPRA algorithm works well with a single label for road networks and max labels of 4-16 for other classes of graphs. The SLPA algorithm performs well with increasing memory size, but overall doesn't offer a favourable return on investment. The RAK algorithm is recommended for label-propagation based disjoint community detection.

show abstract

A review of clique-based overlapping community detection algorithms

Cited by 11 publications

References 87 publications

UCAD: commUnity disCovery method in Attribute-based multicoloreD networks

UCAD: commUnity disCovery method in Attribute-based multicoloreD networks

Detecting Overlapping Communities Based on Influence-Spreading Matrix and Local Maxima of a Quality Function

Selecting a suitable Parallel Label-propagation based algorithm for Disjoint Community Detection

Contact Info

Product

Resources

About