Comparative Deep Models of Language for Minority and Historical Languages
The aim of the Cardamom project is to apply Natural Language Processing and deep learning to a set of minority and historical languages primarily in four language families: Celtic, Germanic, Indic and Dravidian. The methodology involves a big data approach and the creation of single models for closely-related languages, in order to overcome the data gap for under-resourced members within these language families. The results are expected to advance the current state-of-the-art computational models and translate into societal and commercial applications. A more extensive overview can be found under Project description.
The Cardamom project is funded by the Irish Research Council under the Consolidator Laureate Award scheme, with John P. McCrae as principal investigator. The project will run from 2019–2023 and is hosted within the Unit for Linguistic Data in the Insight Centre for Data Analytics, Data Science Institute, National University of Ireland, Galway.