From the extraction of continuous features in parallel texts to visual analytics of heterogeneous areal-typological datasets
2014 (English)In: Language Processing and Grammars: The role of functionally oriented computational models / [ed] Brian Nolan & Carlos Periñán-Pascual, Amsterdam: John Benjamins Publishing Company, 2014, 13-38 p.Chapter in book (Refereed)
The aim of this paper is twofold. First, we show that functionally motivatedprocedural approaches may help to automatically extract typological featuresfrom texts. This idea is illustrated with measuring cross-linguistic variationin the domain of morphological typology based on parallel texts. Second, wedemonstrate that the methodology developed in the field of visual analyticsallows for detecting patterns or regularities in the automatically extractedfeatures. At the heart of our approach lies an extended sunburst visualization,which enables a cross-comparison of a large number of features within thecontext of language genealogy and areal information. We provide evidence of theusefulness of the present approach with case studies where the visualizations ofthe extracted features reveal interesting insights.
Place, publisher, year, edition, pages
Amsterdam: John Benjamins Publishing Company, 2014. 13-38 p.
, Studies in Language Companion Series, ISSN 0165-7763 ; 150
Languages and Literature
IdentifiersURN: urn:nbn:se:su:diva-109336DOI: 10.1075/slcs.150.02mayISBN: 978-90-272-5915-8ISBN: 978-90-272-7064-1OAI: oai:DiVA.org:su-109336DiVA: diva2:764225