DELPH-IN Overview

The DELPH-IN Consortium is a collaboration among computational linguists from research sites world-wide working on ‘deep’ linguistic processing of human language. The goal is the combination of linguistic and statistical processing methods for getting at the meaning of texts and utterances. The partners have adopted Head-Driven Phrase Structure Grammar (HPSG) and Minimal Recursion Semantics (MRS), two advanced models of formal linguistic analysis. They have also committed themselves to a shared format for grammatical representation and to a rigid scheme of evaluation, as well as to the general use of open-source licensing and transparency.

DELPH-IN is a long-standing collaboration over many years involving researchers from the following institutions, among others:

• Bulgarian Academy of Science (Bulgaria), Linguistic Modeling Department
• University of Cambridge (UK), Department of Computer Science and Technology
• Getulio Vargas Foundation (Brazil), Applied Mathematics Department
• IBM Research (Brazil), Brazil Research Lab
• Kyung Hee University (Korea), School of English
• Norwegian University of Science and Technology (Norway), Department of Language and Communication Studies
• Palacký University Olomouc (Czechia), Department of Asian Studies
• CNRS & Université Paris Diderot (France), Laboratoire de Linguistique Formelle (UMR 7110)
• University of Oslo (Norway), Language Technology Group
• University of Sussex (UK), AI Research Group
• University of Washington (US), Computational Linguistics Laboratory
• Vrije Universiteit Amsterdam (Netherlands), Computational Linguistics & Text Mining Lab
• Western Norway University of Applied Sciences (Norway), Department of Language, Literature, Mathematics and Interpreting

Former members

• DFKI Saarbrücken GmbH (Germany), Language Technology Lab (co-founder)
• Melbourne University (Australia), Language Technology Group
• Nanyang Technological University (Singapore), Division of Linguistics and Multilingual Studies
• National Institute of Information and Communications Technology (NICT: Japan)
• NTT Communication Science Laboratories (Japan), Interaction Research Group
• Saarland University (Germany), Department for Computational Linguistics
• Stanford University (US), LinGO Laboratory at CSLI (co-founder)
• Universtitat de Barcelona (Spain), Grup de Recerca Interuniversitari en Aplicacions Lingüístiques (GRIAL)
• University of Lisbon (Portugal), Natural Language and Speech Group
• Universtitat Pompeu Fabra (Spain), Institut Universitari de Lingüística Aplicada (IULA)
• University of Tokyo (Japan)

The DELPH-IN collaboration is open to additional partners and individuals who share our goals and are interested in contributing to the common task.

Components and Resources

Tools and Architectures

• DelphinWelcome
• LKB: Lexical Knowledge Builder — Grammar Engineering Environment
• [incr tsdb()]: Competence and Performance Profiler
• ACE: Answer Constraint Engine, parsing and generation with DELPH-IN grammars
• PyDelphin: Python library for working with DELPH-IN representations
• LOGON: Information about the LOGON machine translation infrastructure.
• Pet: Platform for Experimentation with efficient HPSG processing Techniques
• Other tools: Supporting software, addons, peripheral contributions

Grammars, Frameworks and Treebanks

• English Resource Grammar: DELPH-IN’s most comprehensive grammar.
  • Tutorial on using the ERG, presented at NAACL 2016
  • Documentation of the ERG’s semantic representations
• Catalogue of Grammars, including broad-coverage grammars for German, Japanese, Norwegian, and Spanish, along with significant grammars for several other languages.
• Matrix: Starter-Kit for rapid prototyping of LKB-compatible precision grammars
• CLIMB: Tools to support grammar development of LKB-compatible precision grammars
• Redwoods: HPSG Treebank Comprised of Analyses from the ERG
• MRS: Minimal Recursion Semantics — Theory and Implementation (including extensions and variants such Robust MRS (RMS), Elementary Dependency Structures (EDS) and Dependency MRS (DMRS))
  • ErgSemantics: Emerging documentation of MRS as used in the ERG
  • RmrsDiscussions: Links to discussions related to MRS at various DELPH-IN events
• Shared Corpora, Treebanks
  • MRS Test Suite: Core Test Suite (~100 sentences)
  • Cathedral and the Bazaar: Parallel Corpus based on an Open Source Essay (~800 sentences)
• Grammar discussions: Discussions for grammar developers (analyses, terminology, harmonization, …)
• DELPH-IN RFCs (Requests For Comments; formal specifications)

Applications

• DELPH-IN Applications

About this Site

This is the on-line wiki forum for DELPH-IN software and resources. It serves to enable both developers and users to incrementally create further documentation and up-to-date information on aspects of installation or usage of DELPH-IN technology. Mostly to enforce some discipline among ourselves, these pages require that users are registered to the wiki server in order to obtain write access. Please create a WikiName for yourself, which may require obtaining a ‘textcha’ to protect against wiki spam; once registered at the wiki, to request write access please contact info at delph-in.net. The developers do hope that active DELPH-IN users will contribute to these pages over time.

Archives

Some information from the earlier years of DELPH-IN collaborations is preserved on the following pages, for historical interest:

• Earlier DELPH-IN news items
• Links to earlier DELPH-IN-related projects
• Earlier overviews of the DELPH-IN consortium

Further Information

There is a collection of DELPH-IN mailing lists to which users can subscribe on-line and browse archives of previous postings through the DELPH-IN mailing list manager. If you click on a list, there is a link to the archive near the top of the page.

There is also an active stack-exchange style forum, using the Discourse platform.

DELPH-IN is organized by a standing commitee, but it is pretty informal.

Last update: 2024-08-28 by Francis Bond [edit]