Banpo
SEOUL | KOREA
APRIL 18-21, 2017
News
Sponsors
Previous Events
Channels

Layer Hierachization in Multiplex Networks for Word Cloud Visualization

  • Benjamin Renoust
    National Institute of Informatics, Tokyo, Japan
  • Haolin Ren
    French National Audiovisual Institute (INA)
  • Guy Melancon
    CNRS UMR 5800 LaBRI University of Bordeaux
  • Marie-luce Viaud
    French National Audiovisual Institute (INA)

Abstract

Multiplex networks are heterogeneous and complex data structures: they present interacting entities across multiple layers of interactions. One successful strategy to visualize these complex structures is to exploit an additional network: This layer interaction network describes how layers overlap and may be used to investigate the original network. When the layers correspond to semantic units in documents, the new network shows some sense of semantic hierarchy reflecting the semantic context formed by the documents. We first propose an algorithm to extract hierarchical clusters of layers in a multiplex network of documents, and then to embed this hierarchy in an interactive word cloud.