Improving the use of electronic collocation resources by visual analytics techniques

Authors: Roberto Carlini, Joan Codina-Filba, Leo Wanner

With the increasing prominence of the electronic medium in lexicography, the face of collocation resources also changed. Collocation dictionaries have been extended by additional material (e.g., examples from a corpus and interfaces for targeted access to information), and tools such as Sketch Engine have been developed, which query a corpus and display the collocational (and grammatical) behaviour of a specified word. However, the paradigm of consulting, viewing and exploring the resources still follows to a major extent the traditional dictionary look up philosophy: the user enters a keyword and obtains an outcome in a sequential text format. This implies significant limitations if the user wants to contrast information concerning different keywords or their collocates, view information in incremental detail, etc. Studies on the presentation of information argue that visualization techniques facilitate comprehension. It is thus not by chance that visualization of linguistic information and data has become a popular research topic. In our work, we aim to go one step further: we research how Visual Analytics (VA), which deals with the development of techniques that support the exploration, analysis and interpretation of information, can be used to explore collocation resources in the context of learning Spanish as second language.

Keywords: collocations; active learning; visual analytics

Reference: In Kosem, I., Jakubiček, M., Kallas, J., Krek, S. (eds.) Electronic lexicography in the 21st century: linking lexical data in the digital age. Proceedings of the eLex 2015 conference, 11-13 August 2015, Herstmonceux Castle, United Kingdom. Ljubljana/Brighton: Trojina, Institute for Applied Slovene Studies/Lexical Computing Ltd., pp. 461-476.


Published: 2015