Topic hypergraph: hierarchical visualization of thematic structures in long documents

Guizhen Wang, Chaokai Wen, Binghui Yan, Cong Xie, Ronghua Liang, Wei Chen

Abstract

Thematic information of a long document (i.e., a novel) can be multi-faceted: an interleaving of multiple topics, a sequential evolution of a set of themes, or a crossing superimposition of topics and themes. Conventional topic-based visualization approaches are inefficient to capture this complicated thematic structure. This paper introduces a novel topic-based model, called the topic hypergraph, that characterizes the thematic structure of a long document with a hypergraph representation. Each hypergraph node represents a unique document piece, and encodes its theme as a composition of multiple topics. Two types of relationships among nodes are modeled: an edge that connects two consecutive themes to present their sequential transition, and a hyperedge that encodes a topic. The new representation is essentially a 2D reformulation of the linear streamgraph representation, and can be adaptive by constructing a multi-level hierarchy. We design a suite of visualization and interaction tools to allow users to interactively analyze the theme evolution, theme diversities, and topic interleaving. Our approach is also suitable for comparing multiple long documents.
BibTeX
@article{wang2013topic,
  author = {Guizhen Wang and Chaokai Wen and Binghui Yan and Cong Xie and Ronghua Liang and Wei Chen},
  title = {{Topic hypergraph: hierarchical visualization of thematic structures in long documents}},
  journal = {Science China Information Sciences},
  year = {2013},
  doi = {10.1007/s11432-013-4831-8},
}

Loading related content...

Figure