Elamur Research⌕
☰
home / papers / 2011 / An Empirical Study on Compositionality in Compound Nouns

An Empirical Study on Compositionality in Compound Nouns

Summary

Distributional methods for detecting non-compositional multiword expressions build phrase meaning from word-level semantics. However, earlier evaluation sets either rated whole phrases only or rated constituents only. The authors address this gap for English two-word noun compounds. They built a resource of 90 compounds chosen with WordNet-based heuristics, plus additions from Wiktionary, so that each word could be literal or non-literal. Using Amazon Mechanical Turk, they collected literality ratings for each constituent and for the whole compound. Annotators saw ukWaC example sentences and based their scores on the most frequent sense.

Analysis showed that both nouns strongly shape how compositional a compound is perceived to be. Simple functions combining the two constituent scores closely track the phrase-level ratings. The authors then compared two families of distributional models. Constituent-based models and composition-function models performed comparably, with vector addition slightly ahead. In both families, additive variants beat multiplicative ones. The released data gives a basis for studying how parts contribute to the compositionality of the whole.

Key points

  • The dataset covers 90 compounds, each occurring at least 50 times in ukWaC. Each compound has 3 separate tasks, rated by 30 turkers each, giving 8100 annotations that are publicly released.

  • Spam control combined three steps:

    • a qualification test;
    • filtering by each annotator's average Spearman ρ against the others;
    • rejecting individual ratings more than ±1.5 from the task mean.

    Average inter-annotator ρ was about 0.59.

  • Combining the two constituent literality scores predicts phrase compositionality very well. ADD reaches ρ 0.966 and COMB 0.971, compared with 0.767 and 0.720 for the first or second word alone.

  • In 3-fold cross-validation, the additive composition function av1 + bv2 (a=0.60, b=0.40) scored ρ 0.714. Multiplication scored 0.650, and the best constituent-based model (ADD) scored 0.686. The random baseline was about 0.

Our summary of the paper, in our own words — not its text. The work belongs to its authors; read it at the links above.

Fields: Linguistics & languages

Cited by

100 papers of our library cite it (of the 140673 whose bibliographies we read); 14 of them have a page here, 86 not yet.

Cite

The paper itself, not this page. BibTeX from the library's record of the work.

@inproceedings{reddy2011empirical,
  title = {{An Empirical Study on Compositionality in Compound Nouns}},
  author = {Reddy, Siva and McCarthy, Diana and Manandhar, Suresh},
  booktitle = {Proceedings of 5th International Joint Conference on Natural Language Processing},
  year = {2011},
  url = {https://aclanthology.org/I11-1024/}
}

References

28 references, 7 of them with a card of ours. ↗ leads to the work, «find» searches for one we have no id for yet.

  1. Timothy Baldwin, Colin Bannard, Takaaki Tanaka, and Dominic Widdows. 2003. An empirical model of multiword expression decomposability. In Proceedings of the ACL 2003 workshop on Multiword expressions: analysis, acquisition and treatment, MWE ’03. find our card
  2. Colin Bannard, Timothy Baldwin, and Alex Lascarides. 2003. A statistical approach to the semantics of verb-particles. In Proceedings of the ACL 2003 workshop on Multiword expressions: analysis, acquisition and treatment, MWE ’03. find
  3. Chris Biemann and Eugenie Giesbrecht. 2011. find
  4. Distributional Semantics and Compositionality 2011: Shared Task Description and Results. In Proceedings of DISCo-2011 in conjunction with ACL 2011. find
  5. Paul Cook, Afsaneh Fazly, and Suzanne Stevenson. 2007. Pulling their weight: exploiting syntactic forms for the automatic identification of idiomatic expressions in context. In Proceedings of the Workshop on a Broader Perspective on Multiword Expressions, MWE ’07. find our card
  6. Afsaneh Fazly, Paul Cook, and Suzanne Stevenson. 2009. Unsupervised type and token identification of idiomatic expressions. Computional. Linguististics., 35:61–103, March. find our card
  7. Christiane Fellbaum, editor. 1998. WordNet: An Electronic Lexical Database. The MIT Press. find
  8. A. Ferraresi, E. Zanchetta, M. Baroni, and S. Bernardini. 2008. Introducing and evaluating ukWaC, a very large web-derived corpus of English. In Proceedings of the WAC4 Workshop at LREC 2008. find
  9. Raymond W. Gibbs. 1989. Understanding and Literal Meaning. Cognitive Science, 13(2):243–251. find
  10. Eugenie Giesbrecht. 2009. In Search of Semantic Compositionality in Vector Spaces. In Proceedings of the 17th International Conference on Conceptual Structures: Conceptual Structures: Leveraging Semantic Technologies, ICCS ’09. find
  11. Emiliano Raul Guevara. 2011. Computing Semantic Compositionality in Distributional Semantics. In Proceedings of the Ninth International Conference on Computational Semantics, IWCS ’2011. find
  12. Graham Katz and Eugenie Giesbrecht. 2006. Automatic identification of non-compositional multiword expressions using latent semantic analysis. In Proceedings of the Workshop on Multiword Expressions: Identifying and Exploiting Underlying Properties, MWE ’06. find
  13. Ioannis Korkontzelos and Suresh Manandhar. 2009. Detecting compositionality in multi-word expressions. In Proceedings of the ACL-IJCNLP 2009 Conference Short Papers. find our card
  14. Dekang Lin. 1999. Automatic identification of non-compositional phrases. In Proceedings of the ACL 1999. find our card
  15. Diana McCarthy, Bill Keller, and John Carroll. 2003. Detecting a continuum of compositionality in phrasal verbs. In Proceedings of the ACL 2003 workshop on Multiword expressions: analysis, acquisition and treatment, MWE ’03. find
  16. Diana McCarthy, Sriram Venkatapathy, and Aravind K. Joshi. 2007. Detecting Compositionality of Verb-Object Combinations using Selectional Preferences. In Proceedings of EMNLP-CoNLL 2007. find
  17. Jeff Mitchell and Mirella Lapata. 2008. Vector-based Models of Semantic Composition. In Proceedings of ACL-08: HLT, pages 236–244, Columbus, Ohio. find
  18. Geoffrey Nunberg, Thomas Wasow, and Ivan A. Sag. 1994. Idioms. Language, 70(3):491–539. find
  19. Ted Pedersen. 2011. Identifying Collocations to Measure Compositionality : Shared Task System Description . In Proceedings of DISCo-2011 in conjunction with ACL 2011. find
  20. Siva Reddy, Diana McCarthy, Suresh Manandhar, and Spandana Gella. 2011. Exemplar-Based Word-Space Model for Compositionality Detection: Shared Task System Description. In Proceedings of the ACL Workshop on Distributional Semantics and Compositionality. find
  21. Ivan A. Sag, Timothy Baldwin, Francis Bond, Ann A. Copestake, and Dan Flickinger. 2002. Multiword Expressions: A Pain in the Neck for NLP. In Proceedings of the CICLing 2002. find our card
  22. Magnus Sahlgren. 2006. The Word-Space Model: Using distributional analysis to represent syntagmatic and paradigmatic relations between words in high-dimensional vector spaces. Ph.D. thesis, Stockholm University. find
  23. Patrick Schone and Daniel Jurafsky. 2001. Is Knowledge-Free Induction of Multiword Unit Dictionary Headwords a Solved Problem? In Proceedings of EMNLP 2001. find
  24. Rion Snow, Brendan O’Connor, Daniel Jurafsky, and Andrew Y. Ng. 2008. Cheap and fast—but is it good?: evaluating non-expert annotations for natural language tasks. In Proceedings of EMNLP 2008. find
  25. Caroline Sporleder and Linlin Li. 2009. Unsupervised recognition of literal and non-literal use of idiomatic expressions. In Proceedings of the EACL 2009. find our card
  26. Debra A. Titone and Cynthia M. Connine. 1999. On the compositional and noncompositional nature of idiomatic expressions. Journal of Pragmatics, 31(12):1655 – 1674. Literal and Figurative Language. find
  27. Sriram Venkatapathy and Aravind K. Joshi. 2005. Measuring the relative compositionality of verb-noun (V-N) collocations by integrating features. In Proceedings of the HLT-EMNLP 2005. find
  28. Dominic Widdows. 2008. Semantic Vector Products: Some Initial Investigations. In Second AAAI Symposium on Quantum Interaction, Oxford, March. find

↑ back to top

This page in Markdown → https://md.research.elamur.ai/free/en/papers/reddy-2011-empirical-study-compositionality-compound-nouns/