Elamur Research
☰
home / papers / 2014 / SemLink+: FrameNet, VerbNet and Event Ontologies

SemLink+: FrameNet, VerbNet and Event Ontologies

What it is about

A short review of SemLink, the set of mappings between PropBank, VerbNet, FrameNet and OntoNotes sense groupings, and of the event ontology the authors are building on top of it. The running example is "He talked about politics", labelled Arg0 and Arg1 in PropBank, AGENT and TOPIC in VerbNet, and SPEAKER and TOPIC in FrameNet. The mappings are 1-to-many from PropBank to VerbNet or FrameNet, and mostly 1-to-1 between VerbNet and FrameNet. SemLink 1.2 maps to FrameNet 1.5 frames and frame elements, and to version 3.2 of the PropBank-VerbNet mapping. FrameNet has over 2,000 frame-element types. The paper proposes an event ontology from frame-to-frame relations (Inheritance, Subframe, Using, Causative_of, Precedes) and checks it on the conflict events attack and demonstration. The paper has no experiments and no scores.

Ideas

For Mova

The chapters are languages/en/19-lexicon-mwe-register.md (the lexicon) and languages/en/36-discourse-relations.md (event relations). The code is prag/src/roles.rs and world/src/events.rs. The paper gives no measure, so the first measure is the reproduction of its own examples in the table.

Library summary

The Research library's own summary of the paper, in our words — not the paper's text.

Written for the Research library (claude-opus-5-5) from the paper text, in our own words: the license "CC BY-NC-SA 3.0 (ACL Anthology, before 2016)" does not allow redistributing the text itself. Original — aclanthology.org/W14-3004.pdf

Summary

This workshop paper honors Chuck Fillmore. It argues that FrameNet is central to the problem of event extraction in NLP. Language offers many ways to describe the same event, so systems need to know who did what to whom and how separate events relate. The authors present SemLink, a resource that links PropBank, VerbNet, FrameNet and OntoNotes sense groupings. SemLink connects them in two ways: it maps lexical units across the resources, and it maps the semantic roles of each lexical unit. The goal is to use accurate PropBank-based semantic role labelers to bootstrap VerbNet and FrameNet labels.

The latest release, SemLink 1.2, adds FrameNet frames and Frame Elements and adds OntoNotes groupings to annotated Wall Street Journal data. Its mappings were also checked for consistency.

The authors then turn to relations between events, such as causation, subevents and temporal order. They show how FrameNet's frame-to-frame relations helped the DEFT Richer Event Descriptions group build a partial upper-level Event Ontology. That ontology reconciles ERE event types with FrameNet. SemLink+ is planned to include this ontology. The paper reports no experiments; its contribution is the resource and the analysis.

Key points

  • SemLink links PropBank, VerbNet (version 3.2), FrameNet (FN version 1.5) and OntoNotes (ON) sense groupings.
    • PropBank-to-VerbNet and PropBank-to-FrameNet mappings are usually one-to-many.
    • VerbNet-to-FrameNet mappings tend to be one-to-one.
    • Example: the single PropBank roleset of hear maps to the VerbNet Discover and See classes and to the FrameNet Hear and Perception_experience frames.
  • The three resources differ in how finely they label roles:
    • PropBank uses up to six numbered arguments plus ArgM adjunct tags.
    • VerbNet uses about 30 thematic roles.
    • FrameNet has over 2000 frame-specific Frame Element types.
  • SemLink 1.1 provided the cross-resource mappings and PropBank WSJ instances tagged with VerbNet classes and roles. SemLink 1.2 adds FrameNet frames and Frame Elements and ON groupings. The resulting corpus can be used to train VerbNet class and FrameNet frame classifiers.
  • FrameNet's frame relations proved useful for event modeling. These include Inheritance, Subframe, Perspective_on, Using, Causative_of, Inchoative_of and Precedes. For example, in ERE, attacks and demonstrations are siblings under Conflict events, while FrameNet places them under Hostile_encounter and Protest.
  • The partial Event Ontology combines ERE, FrameNet, SUMO, the Predicate Matrix, WordNet and VerbNet. Work on ERE Life events, such as birth, death and marriage, is ongoing.

Topics: Semantic role labeling

Cite

The paper itself, not this card. BibTeX from the library's record of the work. DOI 10.3115/v1/w14-3004

@inproceedings{palmer2014semlink,
  title = {{SemLink+: FrameNet, VerbNet and Event Ontologies}},
  author = {Palmer, Martha and Bonial, Claire and McCarthy, Diana},
  booktitle = {Proceedings of Frame Semantics in NLP: A Workshop in Honor of Chuck Fillmore (1929-2014)},
  year = {2014},
  doi = {10.3115/v1/w14-3004},
  url = {https://doi.org/10.3115/v1/w14-3004}
}

References

17 references, 2 of them with a card of ours. ↗ leads to the work, «find» searches for one we have no id for yet.

  1. Daniel Bauer & Owen Rambow, 2011, Increasing Coverage of Syntactic Subcategorization Patterns in FrameNet Using VerbNet, In the Proceedings of the IEEE Fifth International Conference on Semantic Computing. find
  2. Claire Bonial, Julia Bonn, Kathryn Conger, Jena Hwang and Martha Palmer, 2014. PropBank: Semantics of New Predicate Types. The 9th edition of the Language Resources and Evaluation Conference. Reykjavik, Iceland. find our card
  3. Claire Bonial, Kevin, Stowe, and Martha Palmer, 2013. Renewing and Revising SemLink. The GenLex Workshop on Linked Data in Linguistics, held with GenLex-13. find
  4. Dipanjan Das, Nathan Schneider, Desai Chen, and Noah A. Smith, 2010. Probabilistic Frame-Semantic Parsing. In Proceedings of the NAACL 2010. find
  5. David Dowty, 1991. Thematic Proto-Roles and Argument Selection. Language, 67:547-619 find
  6. Christiane Fellbaum, 1998. WordNet: An Electronic Lexical Data-base. Language, Speech and Communications. MIT Press Charles. find
  7. J. Fillmore; Collin F. Baker, and H. Sato, 2004. FrameNet as a ``Net". In Proceedings of LREC 2004, 4, pages 1091-1094 find
  8. Charles J. Fillmore, Russell R. Lee-Goldman, and Russell Rhodes. 2012. “The FrameNet Constructicon” Boas, H.C. and Sag, I.A. (Eds.) Sign-based Construction Grammar, CSLI Publications. find
  9. Ana-Maria Guiglea and Alessandro Moschitti. 2006. Semantic role labeling via FrameNet, VerbNet and PropBank. In Proceedings of Coling-ACL 2006, pages 929–936. find
  10. Rei Ikuta and Martha Palmer (2014) Challenges of Adding Causation to Richer Event Descriptions, In the Proceedings of 2nd Events Workshop, held with ACL 2014, Baltimore, MD. find
  11. Karin Kipper, Anna Korhonen, Neville Ryant and Martha Palmer. 2008. A Large-Scale Classification of English Verbs. Language Resources and Evaluation Journal, 42(1):21–40 find
  12. Maddalen Lopez de Lacalle, Egoitz Laparra, German Rigau, 2014, Predicate Matrix: extending SemLink throughWordNet mappings, The 9th edition of the Language Resources and Evaluation Conference. Reykjavik, Iceland. find
  13. Ian Niles and Adam Pease, 2001. Towards a Standard Upper Ontology. In Proceedings of the 2nd International Conference on Formal Ontology in Information Systems (FOIS-2001), Chris Welty and Barry Smith, eds, Ogunquit, Maine, October 17-19, 2001. find
  14. Paola Merlo and Lonneke van der Plas. 2009. Abstraction and generalization in semantic role labels: PropBank, VerbNet or both?, In the Proceedings of ACL 2009. find
  15. Martha Palmer, 2009. SemLink: Linking PropBank, VerbNet and FrameNet, In the Proceedings of the Generative Lexicon Conference, GenLex-09. find
  16. Martha Palmer, Daniel Gildea, and Paul Kingsbury. 2005. The Proposition Bank: An Annotated Corpus of Semantic Roles. Computational Linguistics, 31(1):71–106 Martha Palmer, Daniel Gildea and Nianwen Xue. find our card
  17. Semantic Role Labeling. 2010. Synthesis Lectures on Human Language Technology Series, ed. Graeme Hirst, Morgan and Claypoole. Ralph Weischedel, Eduard Hovy, Mitchell Marcus, Martha Palmer, Robert Belvin, Sameer Pradan, Lance Ramshaw and Nianwen Xue. OntoNotes: A Large Training Corpus for Enhanced Processing, included in Part 1 : Data Acquisition and Linguistic Resources of the Handbook of Natural Language Processing and Machine Translation: Global Automatic Language Exploitation Editors: Joseph Olive, Caitlin Christianson, John McCary, Springer Verglag, pp 54-63, 201 find