Subspace based network community detection using sparse linear coding

Arif Mahmood, Michael Small

    Research output: Contribution to conferenceAbstract

    5 Citations (Scopus)

    Abstract

    © 2016 IEEE.Information mining from networks by identifying communities is an important problem across a number of research fields including social science, biology, physics, and medicine. Most existing community detection algorithms are graph theoretic and lack the ability to detect accurate community boundaries if the ratio of intra-community to inter-community links is low. Also, algorithms based on modularity maximization may fail to resolve communities smaller than a specific size if the community size varies significantly. We propose a fundamentally different community detection algorithm based on the fact that each network community spans a different subspace in the geodesic space. Therefore, each node can only be efficiently represented as a linear combination of nodes spanning the same subspace (Fig. 1). To make the process of community detection more robust, we use sparse linear coding with l1 norm constraint. In order to find a community label for each node, sparse spectral clustering algorithm is used. The proposed community detection technique is compared with more than ten state of the art methods on two benchmark networks (with known clusters) using normalized mutual information criterion. Our proposed algorithm outperformed existing methods with a significant margin on both benchmark networks.
    Original languageEnglish
    Pages1502-1503
    Number of pages2
    DOIs
    Publication statusPublished - 2016
    Event2015 Ecological Society of Australia Conference - Adelaide, Australia
    Duration: 29 Nov 20153 Dec 2015

    Conference

    Conference2015 Ecological Society of Australia Conference
    CountryAustralia
    CityAdelaide
    Period29/11/153/12/15

    Fingerprint

    Social sciences
    Clustering algorithms
    Medicine
    Labels
    Physics

    Cite this

    Mahmood, A., & Small, M. (2016). Subspace based network community detection using sparse linear coding. 1502-1503. Abstract from 2015 Ecological Society of Australia Conference, Adelaide, Australia. https://doi.org/10.1109/ICDE.2016.7498395
    Mahmood, Arif ; Small, Michael. / Subspace based network community detection using sparse linear coding. Abstract from 2015 Ecological Society of Australia Conference, Adelaide, Australia.2 p.
    @conference{de4702d04d474ed1893e70a616deada9,
    title = "Subspace based network community detection using sparse linear coding",
    abstract = "{\circledC} 2016 IEEE.Information mining from networks by identifying communities is an important problem across a number of research fields including social science, biology, physics, and medicine. Most existing community detection algorithms are graph theoretic and lack the ability to detect accurate community boundaries if the ratio of intra-community to inter-community links is low. Also, algorithms based on modularity maximization may fail to resolve communities smaller than a specific size if the community size varies significantly. We propose a fundamentally different community detection algorithm based on the fact that each network community spans a different subspace in the geodesic space. Therefore, each node can only be efficiently represented as a linear combination of nodes spanning the same subspace (Fig. 1). To make the process of community detection more robust, we use sparse linear coding with l1 norm constraint. In order to find a community label for each node, sparse spectral clustering algorithm is used. The proposed community detection technique is compared with more than ten state of the art methods on two benchmark networks (with known clusters) using normalized mutual information criterion. Our proposed algorithm outperformed existing methods with a significant margin on both benchmark networks.",
    author = "Arif Mahmood and Michael Small",
    year = "2016",
    doi = "10.1109/ICDE.2016.7498395",
    language = "English",
    pages = "1502--1503",
    note = "2015 Ecological Society of Australia Conference ; Conference date: 29-11-2015 Through 03-12-2015",

    }

    Mahmood, A & Small, M 2016, 'Subspace based network community detection using sparse linear coding' 2015 Ecological Society of Australia Conference, Adelaide, Australia, 29/11/15 - 3/12/15, pp. 1502-1503. https://doi.org/10.1109/ICDE.2016.7498395

    Subspace based network community detection using sparse linear coding. / Mahmood, Arif; Small, Michael.

    2016. 1502-1503 Abstract from 2015 Ecological Society of Australia Conference, Adelaide, Australia.

    Research output: Contribution to conferenceAbstract

    TY - CONF

    T1 - Subspace based network community detection using sparse linear coding

    AU - Mahmood, Arif

    AU - Small, Michael

    PY - 2016

    Y1 - 2016

    N2 - © 2016 IEEE.Information mining from networks by identifying communities is an important problem across a number of research fields including social science, biology, physics, and medicine. Most existing community detection algorithms are graph theoretic and lack the ability to detect accurate community boundaries if the ratio of intra-community to inter-community links is low. Also, algorithms based on modularity maximization may fail to resolve communities smaller than a specific size if the community size varies significantly. We propose a fundamentally different community detection algorithm based on the fact that each network community spans a different subspace in the geodesic space. Therefore, each node can only be efficiently represented as a linear combination of nodes spanning the same subspace (Fig. 1). To make the process of community detection more robust, we use sparse linear coding with l1 norm constraint. In order to find a community label for each node, sparse spectral clustering algorithm is used. The proposed community detection technique is compared with more than ten state of the art methods on two benchmark networks (with known clusters) using normalized mutual information criterion. Our proposed algorithm outperformed existing methods with a significant margin on both benchmark networks.

    AB - © 2016 IEEE.Information mining from networks by identifying communities is an important problem across a number of research fields including social science, biology, physics, and medicine. Most existing community detection algorithms are graph theoretic and lack the ability to detect accurate community boundaries if the ratio of intra-community to inter-community links is low. Also, algorithms based on modularity maximization may fail to resolve communities smaller than a specific size if the community size varies significantly. We propose a fundamentally different community detection algorithm based on the fact that each network community spans a different subspace in the geodesic space. Therefore, each node can only be efficiently represented as a linear combination of nodes spanning the same subspace (Fig. 1). To make the process of community detection more robust, we use sparse linear coding with l1 norm constraint. In order to find a community label for each node, sparse spectral clustering algorithm is used. The proposed community detection technique is compared with more than ten state of the art methods on two benchmark networks (with known clusters) using normalized mutual information criterion. Our proposed algorithm outperformed existing methods with a significant margin on both benchmark networks.

    U2 - 10.1109/ICDE.2016.7498395

    DO - 10.1109/ICDE.2016.7498395

    M3 - Abstract

    SP - 1502

    EP - 1503

    ER -

    Mahmood A, Small M. Subspace based network community detection using sparse linear coding. 2016. Abstract from 2015 Ecological Society of Australia Conference, Adelaide, Australia. https://doi.org/10.1109/ICDE.2016.7498395