Apart from the traditional field of corpus-oriented tagging and parsing, some of the most dynamic fields in CG research right now are arguably machine translation and tools for less-resourced languages. Thus, in addition to ongoing work at GramTrans, the open source MT initiative Apertium has begun to use Constraint Grammar, and CG is flourishing for many minor languages, such as the Sami languages, Greenlandic, Basque, Tibetan and the Celtic languages. Both of these fields share a need for high-quality lexical, morphological and semantic resources as input to CG grammars and applications. We therefore also invited contributions concerning research in fields relevant to the CG framework on the input side, such as finite-state analyzers, ontologies etc. Finally, we hoped for methodological contributions, exploiting advances in expressive power in the most widely used CG compiler, CG3.
All in all, we thought that there was a sound basis for a workshop in the area, and our hopes were confirmed by high quality of the submitted papers, whose number almost became a problem for the planned time frame of a half-day workshop. Encouragingly, the workshop contributions covered as many languages as there were papers (7), with forages into very diverse language families. Although the expected focus on machine translation did not manifest itself this time, the range of topics was considerable and included both low and high levels of linguistic analysis (morphology, syntax, anaphora), and - on the theoretical side - research on both mathematical, methodological and machine learning issues (SAT, FST and grammar optimization).
We would like to thank the NoDaLiDa organizers for an inspiring and efficient cooperation, and our program committee for their thorough and expert work in the triple reviewing of all submitted papers.
On behalf of the workshop organizers
Eckhard Bick and Kristin Hagen