Visualising a Text with a Tree Cloud

Philippe Gambette (1) , Jean Véronis (2)

1 ALGCO - Algorithmes, Graphes et Combinatoire
2 LIF - Laboratoire d'informatique Fondamentale de Marseille - UMR 6166

Tag clouds have gained popularity over the internet to provide a quick overview of the content of a website or a text. We introduce a new visualisation which displays more information: the tree cloud. Like a word cloud, it shows the most frequent words of the text, where the size reflects the frequency, but the words are arranged on a tree to reflect their semantic proximity according to the text. Such tree clouds help identify the main topics of a document, and even be used for text analysis. We also provide methods to evaluate the quality of the obtained tree cloud, and some key steps of its construction. Our algorithms are implemented in the free software TreeCloud available at http://www.treecloud.org

Keywords: Information visualisation; tag cloud; semantic proximity; hierarchical clustering; arboricity

Domains: Document and Text Processing; Computation and Language [cs.CL]

Philippe Gambette, Jean Véronis. Visualising a Text with a Tree Cloud. IFCS'09: International Federation of Classification Societies Conference, Mar 2009, Dresde, Germany. pp.561-569, ⟨10.1007/978-3-642-10745-0_61⟩. ⟨lirmm-00373643v2⟩