Quantifying linguistic data


On 21 November, Professor Pascual Cantos Gómez conducted a four-hour seminar on "Quantifying linguistic data: statistical description, decision taking and modeling". After completing a Masters programme in computational linguistics in Essex, Prof Cantos got his PhD at the University of Murcia, where he is Full Professor in English linguistics and member of the LACELL (Linguistics Applied to Computation, Language Learning and Lexicography) research group. His research fields are corpus and computational linguistics, quantitative linguistics, lexicography and computer-assisted language learning (CALL). Prof Castos has published extensively in such areas, in chapters and monographs published by Equinox, John Benjamins, Springer, Lancaster University Press and Elsevier, among others. In this respect, it is worth mentioning his 2011 monograph (published in 2013 by Equinox) Statistical methods in language and linguistic research. As regards his journal articles, Pascual Cantos has published in International Journal of Corpus Linguistics, Computer and the Humanities, etc.

Prof Cantos’s seminar in Vigo can be summarised as follows. Statistics allows a quantitative approach to linguistic research. However, most of the research in the field of language and linguistics is qualitative. In this seminar, Pascual Cantos used both approaches and presented those quantitative methods and statistical techniques that can be applied to linguistic research in order to sustain and guarantee better data, better analysis and better qualitative results. To do this, he introduced the basic descriptive metrics and later dealt with some parametric and non-parametric techniques. Prof Cantos also described some mathematical properties of natural language modeling and decision. The dynamics of the seminar were highly practical. The lecturer started with brief explanations of some basic theoretical concepts and immediately suggested  Excel- and R-based hands-ons.