A distributed parallel community detection heuristics for large-scale real graphs (Unrefereed Workshop Manuscript)

• Item type: SIG Technical Reports(1)
• 公開日: 2015-07-28
• タイトル: A distributed parallel community detection heuristics for large-scale real graphs (Unrefereed Workshop Manuscript)
• タイトル / 言語: en
• タイトル: A distributed parallel community detection heuristics for large-scale real graphs (Unrefereed Workshop Manuscript)
• 言語: eng
• キーワード / 主題Scheme: Other
• キーワード / 主題: ビッグデータ
• 資源タイプ / 資源タイプ識別子: http://purl.org/coar/resource_type/c_18gh
• 資源タイプ: technical report
• 著者所属: Department of Human System Science／Presently with Tokyo Institute of Technology
• 著者所属: Department of Mathematical and Computing Sciences／Presently with Tokyo Institute of Technology
• 著者所属: Global Scientific Information and Computing Center／Presently with Tokyo Institute of Technology
• 著者所属(英): en
• 著者所属(英): Department of Human System Science / Presently with Tokyo Institute of Technology
• 著者所属(英): en
• 著者所属(英): Department of Mathematical and Computing Sciences / Presently with Tokyo Institute of Technology
• 著者所属(英): en
• 著者所属(英): Global Scientific Information and Computing Center / Presently with Tokyo Institute of Technology
• 著者名: Guanglong, Chi; Ken, Wakita; Hitoshi, Sato
• 著者名(英): Guanglong, Chi; Ken, Wakita; Hitoshi, Sato
• 論文抄録 / 内容記述タイプ: Other
• 論文抄録 / 内容記述: In the era of ”Big Data”, the need for fast, scalable, and high-precision technique for analysis of huge social networks is growing. The paper proposes a distributed implementation scheme of a community structure identification technique known as the Louvain method [5]. Our technique copes with the growing size of the social network by partitioning and distributing the whole social network. It also employs a few approximation techniques and collective communication to increase parallelisms among the computing nodes. The performance is evaluated using various social network data as well as synthesized data.
• 論文抄録(英) / 内容記述タイプ: Other
• 論文抄録(英) / 内容記述: In the era of ”Big Data”, the need for fast, scalable, and high-precision technique for analysis of huge social networks is growing. The paper proposes a distributed implementation scheme of a community structure identification technique known as the Louvain method [5]. Our technique copes with the growing size of the social network by partitioning and distributing the whole social network. It also employs a few approximation techniques and collective communication to increase parallelisms among the computing nodes. The performance is evaluated using various social network data as well as synthesized data.
• 書誌レコードID / 収録物識別子タイプ: NCID
• 書誌レコードID / 収録物識別子: AN10463942
• 書誌情報: 研究報告ハイパフォーマンスコンピューティング（HPC）; 巻 2015-HPC-150, 号 22, p. 1-11, 発行日 2015-07-28
• ISSN / 収録物識別子タイプ: ISSN
• ISSN / 収録物識別子: 2188-8841
• Notice: SIG Technical Reports are nonrefereed and hence may later appear in any journals, conferences, symposia, etc.
• 出版者 / 言語: ja
• 出版者: 情報処理学会