May 9, 2016
A workshop presenting the final outcomes of the Welsh Natural Language Toolkit Project funded by the Welsh-language Technology and Digital Media Grant (Andreas Vlachidis, Daniel Cunliffe, Douglas Tudhope) was held 28th April at GBC, Trefforest.
The project developed a suite of open source Natural Language Processing modules for the Welsh language. Participation included Bangor University, Cardiff University, Royal Commission on the Ancient and Historical Monuments of Wales, Sheffield University and the Welsh Government.
Daniel introduced the project and its background. Andreas outlined the rationale of the work revealing the methods and techniques used for building a range of core NLP modules for Welsh. The presentation discussed how the project has tacked linguistic behaviours of Welsh relating to use of hyphenation and apostrophe, formation of plural nouns, part of speech tagging and contact mutation, etc. Discussion ranged over a number of different applications of the toolkit for Welsh language text.
Presentations were also given by Gareth Morlais (Welsh Government) and Angus Roberts from the GATE team (University of Sheffield).