PARSing and Multi-word Expressions

EU IC1207 COST Action project – aims to improve accuracy of natural language parsing by investigating and focusing on multi word expressions. The project is carried out by researchers from 30 countries and results will be applied to more than 29 languages and 6 dialects from 10 language families.


The IC1207 COST ActionPARSEME, is an interdisciplinary scientific network devoted to the role of multi-word expressions (MWEs) in parsing.

  • It gathers interdisciplinary experts (linguists, computational linguists, computer scientists, psycholinguists, and industrials) from 30 countries which have signed the Memorandum of Understanding.
  • It represents 29 languages and 6 dialects from 10 language families1.
  • It covers different parsing  frameworks: CCG (Combinatory Categorial Grammar), DG (Dependancy Grammar), GG (Generative Grammar), HPSG (Head-driven Phrase Structure Grammar), LFG (Lexical Functional Grammar), TAG (Tree Adjoining Grammar), …
  • It addresses different methodologies (symbolic, probabilistic and hybrid parsing) and language technology applications (machine translation, information retrieval, …).

PARSEME is structured in 4 Working Groups. Its main activities include scientific meetingsshort-term scientific missionsworkshops and training schools. It promotes gender balance and it adheres to COST policy in promoting early-stage researchers.
Its taget groups are the Action’s members, early-stage researchers, Master and PhD students, related Actions, projects and networks, other NLP researchers, language industry professionals, language resource providers, European institutions, foreign language teachers and learners, as well as the general public.

PARSEME welcomes new members from any member countries or from new COST countries.

Project fact sheet

Participants: Numerous researchers from different scientific disciplines – TakeLab’s Jan Šnajder is one of the participants
Duration: 4 years (March 8, 2013 – March 7, 2017)