This page PARSEME deliverables, as defined in the Memorandum of Understanding.

  1. Contrastive analysis of the linguistic properties of MWEs in different European languages.
  2. Proposal of a common design for lexicons including both valence data and MWE data.
    • Publication on contrastive analysis of the design of valence MWE dictionaries in Czech and Polish:
      Adam Przepiórkowski, Jan Hajič, Elżbieta Hajnicz, and Zdeňka Urešová. Phraseology in two Slavic valency dictionaries: Limitations and perspectives. International Journal of Lexicography, 30(1):1–38, 2017
    • WG1 workshop on lexical encoding of MWEs, based on the DuELME formalism meant to be theory- and grammar-independent, and interoperable with valence-aware grammars
  3. Lexical databases: possibly interoperable parsing-oriented MWE lexicons and valence dictionaries in several European languages.
  4. Extensions of existing corpora and treebanks in several languages with MWE annotation levels.
    • Annotation guidelines for 18 languages in the PARSEME shared task on automatic identification of verbal MWEs
    • Course on MWEs and the Praque Dependency Treebank
    • PARSEME-FR - a PARSEME French spin-off project with annotating MWEs as one of the main objectives
    • Papers on projecting MWE resources on treebanks
  5. Extensions of existing grammars for several European languages with rules dedicated to MWEs.
    • Course on Multi-word Expressions in HPSG
    • Posters and papers on integrating MWEs in symbolic grammars (in English, Greek, Hebrew and Polish)
  6. Definitions of abstract models (e.g. meta-grammars) of MWEs’ properties that would: (i) capture linguistic richness of MWEs independently of particular grammatical frameworks, (ii) help reduce the cost of resource development, (iii) adapt to different languages studied.
    • 2 tutorials on XMG, a meta-grammar framework for efficient development of lexicalized grammars with MWEs
    • Papers and posters on MWE encoding in XMG
    • A tutorial on integrating MWEs in FRMG, a French Meta-Grammar
    • WG1 workshop on lexical encoding of MWEs, based on the DuELME formalism meant to be theory- and grammar-independent, and interoperable with valence-aware grammars
  7. Recommendations of best practices for MWE representation and treatment in parsing within different theoretical frameworks.
    • Papers on joint parsing and MWE identification
    • Tutorials on MWEs in FRMG and in the Grammatical Framework
    • Course on "Dependency grammar, dependency parsing and MWEs"
    • WG2 book "Representation and Parsing of Multiword Expressions"
  8. Extension of hybrid (knowledge-based and data-driven) methods for parsing MWEs.
    • WG3 survey on hybrid processing of MWEs
    • Papers on a novel architecture of joint dependency parsing and MWE identification
    • Papers on promoting MWEs in TAG parsing
  9. Annotation guidelines for the representation of MWEs in treebanks.
    • WG4 survey on annotating MWEs in treebanks
    • 2 papers describing the survey and paving the way towards guidelines
  10. A common publishing platform gathering initiatives in the field of MWEs and parsing.
    • This website
    • Publicly available Google table from the WG1 MWE lexicon survey
    • Publicly available Wiki table from the WG4 survey on annotating MWEs in treebanks
  11. Scientific publications in established conferences and journals in various domains - see the pages dedicated to papers and proceedings.