Title

A Step Forward to Hypertext

Authors

Adán Cassán (Department of Computational Linguistics, Planeta Actimedia C/ Aribau, 198, 5ª planta, 08036, Barcelona, Spain)

Sergi Cervell (Department of Computational Linguistics, Planeta Actimedia C/ Aribau, 198, 5ª planta, 08036, Barcelona, Spain)

Mireia Colom (Department of Computational Linguistics, Planeta Actimedia C/ Aribau, 198, 5ª planta, 08036, Barcelona, Spain)

Rafael Marín (Department of Computational Linguistics, Planeta Actimedia C/ Aribau, 198, 5ª planta, 08036, Barcelona, Spain)

Josep M. Merenciano (Department of Computational Linguistics, Planeta Actimedia C/ Aribau, 198, 5ª planta, 08036, Barcelona, Spain)

Gema Pérez (Department of Computational Linguistics, Planeta Actimedia C/ Aribau, 198, 5ª planta, 08036, Barcelona, Spain)

Lluís Valentín (Department of Computational Linguistics, Planeta Actimedia C/ Aribau, 198, 5ª planta, 08036, Barcelona, Spain)

Session

WO23: Corpus Analysis, Annotation, Representation

Abstract

In this paper, after a critical review of how hypertext has been understood over the past few years, we claim against the distinction between total and partial hypertext, and we provide a brief description of a dynamic system that allows the automatic highlighting of those textual elements related to a certain topic. The outcome of our approach is ESQUITX, an automatic highlighter based on different filters, particularly those referring to topic information. The general process can be summarized as follows: once the text is lemmatized, by means of our Spanish tagger, a collection of filters is applied, and only the resulting lemma forms are highlighted. 

Keywords

Hypertext

Full Paper

268.pdf