• DocumentCode
    1577958
  • Title

    Analysis and Visualization of Dynamic Clusterings

  • Author

    Held, Pascal ; Kruse, Rudolf

  • Author_Institution
    Dept. of Comput. Sci., Otto von Guericke Univ., Magdeburg, Germany
  • fYear
    2013
  • Firstpage
    1385
  • Lastpage
    1393
  • Abstract
    Clusterings in dynamic networks are also dynamic. This means that they change over time. In this paper we present a visualization to show these changing behavior. For this purpose we used and modified the MONIC framework to track the clusters over time. Possible transactions during the lifetime of a cluster are birth, death, growth, contraction, splitting, and merging. We extend this list by rebirth, for clusters which were reactivated after death. In our evolution diagram we used lines to represent the lifetime of a cluster. Splitting and merging clusters have connecting lines. For the line adjustment we used a bandwidth reduction on connected components.
  • Keywords
    data analysis; data visualisation; graph theory; pattern clustering; MONIC framework; bandwidth reduction; birth; connecting lines; contraction; death; dynamic clustering analysis; dynamic clustering visualization; dynamic graphs; growth; huge data set analysis; merging clusters; splitting clusters; Algorithm design and analysis; Clustering algorithms; Data visualization; Heuristic algorithms; Merging; Monitoring; Postal services;
  • fLanguage
    English
  • Publisher
    ieee
  • Conference_Titel
    System Sciences (HICSS), 2013 46th Hawaii International Conference on
  • Conference_Location
    Wailea, Maui, HI
  • ISSN
    1530-1605
  • Print_ISBN
    978-1-4673-5933-7
  • Electronic_ISBN
    1530-1605
  • Type

    conf

  • DOI
    10.1109/HICSS.2013.93
  • Filename
    6480003