Elamur Research⌕
☰
home / papers / 2013 / Automatic Interpretation of the English Possessive

Automatic Interpretation of the English Possessive

Paper · 2013 · Stephen Tratz and Eduard Hovy

Summary

Tratz and Hovy address a gap in computational semantics: the English ’s construction (with possessive pronouns counted as the same phenomenon) is frequent and expresses many relations, such as ownership, kinship, part-whole, time and measure, yet had seen little NLP work and no public data. They built an inventory of 17 relations plus OTHER. It was formed by grouping real corpus examples and then refined over several annotation rounds, with approximate mappings to the schemes of Quirk et al. and Badulescu and Moldovan. They annotated 21,938 possessives from three genres: Penn Treebank newswire, Gibbon's history of Rome, and Kipling's The Jungle Book. A final three-annotator study, which included a constrained revision step, reached Fleiss’ Kappa 0.78. An SVM classifier using lexical, WordNet-derived and syntactic-context features reached 87.4% accuracy on held-out data, the best reported figure for this task at the time. The dataset and software were released publicly. This gives later systems both a usable label set and training material for interpreting possessives.

Key points

  • Dataset size and sources: 21,938 examples in total.
    • 15,330 come from Penn Treebank sections 2–21, extracted from gold parse trees.
    • Gibbon's multi-volume account of the Roman Empire's collapse supplied 5,266, and The Jungle Book supplied 1,342. Both were parsed automatically with the authors' own parser.
  • Frequency: ’s occurs roughly 1.8 times per 100 words in the Penn Treebank.
  • Agreement study: a final set of 150 examples (50 per source) was labelled by three annotators.
    • Full three-way agreement rose from 82 to 109 examples after a revision pass.
    • Fleiss’ Kappa rose from 0.60 to 0.78.
  • Comparison with prior work: Badulescu and Moldovan had only 960 ’s examples and did not release their data. Their best system scored F 78.75, and their SVM managed just 23.25%.
  • Feature ablation (leave-one-out vs only-one):
    • Dropping any single feature template barely changes accuracy (about 0.867–0.875).
    • The strongest template used alone, gloss terms, reaches about 0.76.

Interesting ideas

  • I1. The relation inventory was induced bottom-up by clustering real examples, then explicitly mapped to earlier taxonomies, so each label is grounded in data and comparable to other schemes.
  • I2. Each annotation batch was sampled so that no possessee word repeats, which forces diversity and stops a few frequent heads from dominating agreement estimates.
  • I3. Disagreements were resolved by blind re-annotation limited to the labels already proposed by someone, without showing who chose what or how many did.
  • I4. Pairing leave-one-out with only-one ablation separates redundant features from individually informative ones, since removal alone can hide useful but overlapping features.
  • I5. Annotators could record a second-best label, and the authors plan to output every acceptable interpretation rather than one winner, which treats some ambiguity as real rather than noise.

For Mova

How these ideas could serve Mova, our small open language model — its files, stages and tests:

  • I1 (world): Mova could split nmod:poss into distinct world relations (holding, kinship, part-of, member-of, time, extent) instead of one generic "has". Each relation would link to its WordNet/FrameNet counterpart, so the world model can explain which reading of a possessive it chose and why.
  • I2 (learn): when sampling silver or LLM-teacher items for evaluation, enforce unique head lemmas per batch so that measured accuracy reflects coverage rather than repetition of common cases.
  • I3 (learn): when witness parsers or teacher tiers disagree, an adjudicator could choose only among the candidates they produced, with their sources anonymized, to reduce authority bias in the agreed silver data.
  • I4 (learn): run both ablation styles on the parser's and frame-labeller's feature templates to find which explainable cues truly carry signal and which can be pruned.
  • I5 (frames): frame building could keep a ranked set of possible readings for ambiguous possessives and let later world or answer evidence choose among them, instead of committing early.

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

 Institutions

🎓Carnegie Mellon UniversityPittsburgh · United States539 works · 179,533 citations🔬DEVCOM Army Research LaboratoryAdelphi · United States14 works · 2,292 citations

 Fields, topics and concepts

📚 Shelves 4The field of science this work belongs to (OpenAlex fields).Computer Science · 16,667 worksArtificial Intelligence · 16,194 worksArts and Humanities · 16,039 worksLanguage and Linguistics · 15,891 works

🧭 Topics 1Research subjects and languages the work is about.English · 2,263 works

Cite

Cite as Stephen Tratz and Eduard Hovy. 2013. Automatic Interpretation of the English Possessive. Proceedings of the 51st Annual Meeting of the Association for Computational Linguistics (Volume 1: Long Papers).

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

@inproceedings{tratz2013automatic,
  title = {{Automatic Interpretation of the English Possessive}},
  author = {Stephen Tratz and Eduard Hovy},
  booktitle = {Proceedings of the 51st Annual Meeting of the Association for Computational Linguistics (Volume 1: Long Papers)},
  year = {2013},
  url = {https://aclanthology.org/P13-1037/}
}

References

37 references, 4 of them with a card of ours. ↗ leads to the work, «find» searches for one we have no id for yet.

  1. Adriana Badulescu and Dan Moldovan. 2009. A Semantic Scattering Model for the Automatic Interpretation of English Genitives. Natural Language Engineering, 15:215–239. ↗
  2. Chris Barker. 1995. Possessive Descriptions. CSLI Publications, Stanford, CA, USA. find
  3. Cristina Butnariu and Tony Veale. 2008. A Concept-Centered Approach to Noun-Compound Interpretation. In Proceedings of the 22nd International Conference on Computational Linguistics, pages 81–88. find
  4. Cristina Butnariu, Su Nam Kim, Preslav Nakov, Diarmuid Ó Séaghdha, Stan Szpakowicz, and Tony Veale. 2009. SemEval-2010 Task 9: The Interpretation of Noun Compounds Using Paraphrasing Verbs and Prepositions. In DEW ’09: Proceedings of the Workshop on Semantic Evaluations: Recent Achievements and Future Directions, pages 100– 105. Cristina Butnariu, Su Nam Kim, Preslav Nakov, Diarmuid. find
  5. Ó Séaghdha, Stan Szpakowicz, and Tony Veale. 2010. SemEval-2010 Task 9: The Interpretation of Noun Compounds Using Paraphrasing Verbs and Prepositions. In Proceedings of the 5th International Workshop on Semantic Evaluation, pages 39–44. find
  6. Jacob Cohen. 1960. A Coefficient of Agreement for Nominal Scales. Educational and Psychological Measurement, 20(1):37–46. ↗
  7. Rong-En Fan, Kai-Wei Chang, Cho-Jui Hsieh, Xiang-Rui Wang, and Chih-Jen Lin. 2008. LIBLINEAR: A Library for Large Linear Classification. Journal of Machine Learning Research, 9:1871–1874. ↗
  8. Christiane Fellbaum. 1998. WordNet: An Electronic Lexical Database. The MIT Press, Cambridge, MA. find
  9. Joseph L. Fleiss. 1971. Measuring nominal scale agreement among many raters. Psychological Bulletin, 76(5):378–382. Edward Gibbon. 1776. The History of the Decline and Fall of the Roman Empire, volume I of The History of the Decline and Fall of the Roman Empire. Printed for W. Strahan and T. Cadell. ↗
  10. Roxanna Girju. 2009. The Syntax and Semantics of Prepositions in the Task of Automatic Interpretation of Nominal Phrases and Compounds: A Cross-linguistic Study. Computational Linguistics - Special Issue on Prepositions in Application, 35(2):185–228. ↗
  11. Bernd Heine. 1997. Possession: Cognitive Sources, Forces, and Grammaticalization. Cambridge University Press, United Kingdom. find
  12. Iris Hendrickx, Su Nam Kim, Zornitsa Kozareva, Preslav Nakov, Diarmuid Ó Séaghdha, Sebastian Padó, Marco Pennacchiotti, Lorenza Romano, and Stan Szpakowicz. 2009. Semeval-2010 task 8: Multi-Way Classification of Semantic Relations between Pairs of Nominals. In Proceedings of the Workshop on Semantic Evaluations: Recent Achievements and Future Directions (SEW-2009), pages 94–99.
  13. Iris Hendrickx, Zornitsa Kozareva, Preslav Nakov, Diarmuid Ó Séaghdha, Stan Szpakowicz, and Tony Veale. 2013. Task Description: SemEval-2013 Task 4: Free Paraphrases of Noun Compounds. http://www.cs.york.ac.uk/semeval-2013/task4/. [Online; accessed 1-May-2013]. ↗
  14. Su Nam Kim and Timothy Baldwin. 2005. Automatic Interpretation of Noun Compounds using Word-Net::Similarity. Natural Language Processing– IJCNLP 2005, pages 945–956. find
  15. Rudyard Kipling. 1894. The Jungle Book. Macmillan, London, UK. find
  16. George Lakoff and Mark Johnson. 1980. Metaphors We Live by. The University of Chicago Press, Chicago, USA. find
  17. George Lakoff. 1987. Women, Fire, and Dangerous Things: What Categories Reveal about the Mind. The University of Chicago Press, Chicago, USA. find
  18. Mark Lauer. 1995. Corpus Statistics Meet the Noun Compound: Some Empirical Results. In Proceedings of the 33rd Annual Meeting on Association for Computational Linguistics, pages 47–54. ↗
  19. Ken Litkowski and Orin Hargraves. 2007. SemEval-2007 Task 06: Word-Sense Disambiguation of Prepositions. In Proceedings of the 4th International Workshop on Semantic Evaluations, pages 24–29. find
  20. Sebastian Löbner. 1985. Definites. Journal of Semantics, 4(4):279. find
  21. Mitchell P. Marcus, Mary A. Marcinkiewicz, and Beatrice Santorini. 1993. Building a Large Annotated Corpus of English: The Penn Treebank. Computational Linguistics, 19(2):330. ↗
  22. Adam Meyers, Ruth Reeves, Catherine Macleod, Rachel Szekely, Veronika Zielinska, Brian Young, and Ralph Grishman. 2004. The NomBank Project: An Interim Report. In Proceedings of the NAACL/HLT Workshop on Frontiers in Corpus Annotation.
  23. Dan Moldovan and Adriana Badulescu. 2005. A Semantic Scattering Model for the Automatic Interpretation of Genitives. In Proceedings of Human Language Technology Conference and Conference on Empirical Methods in Natural Language Processing, pages 891–898. find
  24. Preslav Nakov. 2008. Noun Compound Interpretation Using Paraphrasing Verbs: Feasibility Study. In Proceedings of the 13th International Conference on Artificial Intelligence: Methodology, Systems, and Applications, pages 103–117. find
  25. Vivi Nastase and Stan Szpakowicz. 2003. Exploring Noun-Modifier Semantic Relations. In Fifth International Workshop on Computational Semantics (IWCS-5), pages 285–301. find
  26. Kiki Nikiforidou. 1991. The Meanings of the Genitive: A Case Study in the Semantic Structure and Semantic Change. Cognitive Linguistics, 2(2):149– 206. find
  27. Sumiyo Nishiguchi. 2009. Qualia-Based Lexical Knowledge for the Disambiguation of the Japanese Postposition No. In Proceedings of the Eighth International Conference on Computational Semantics. find
  28. Diarmuid Ó Séaghdha and Ann Copestake. 2009. Using Lexical and Relational Similarity to Classify Semantic Relations. In Proceedings of the 12th Conference of the European Chapter of the Association for Computational Linguistics, pages 621–629. find
  29. Barbara H. Partee and Vladimir Borschev. 1998. Integrating Lexical and Formal Semantics: Genitives, Relational Nouns, and Type-Shifting. In Proceedings of the Second Tbilisi Symposium on Language, Logic, and Computation, pages 229–241. find
  30. James Pustejovsky. 1995. The Generative Lexicon. MIT Press, Cambridge, MA, USA. ↗
  31. Randolph Quirk, Sidney Greenbaum, Geoffrey Leech, and Jan Svartvik. 1985. A Comprehensive Grammar of the English Language. Longman Inc., New York. find
  32. Anette Rosenbach. 2002. Genitive Variation in English: Conceptual Factors in Synchronic and Diachronic Studies. Topics in English linguistics. Mouton de Gruyter. find
  33. Sidney Siegel and N. John Castellan. 1988. Non-parametric statistics for the behavioral sciences. McGraw-Hill. find
  34. John R. Taylor. 1996. Possessives in English. Oxford University Press, New York. ↗
  35. Stephen Tratz and Eduard Hovy. 2010. A Taxonomy, Dataset, and Classifier for Automatic Noun Compound Interpretation. In Proceedings of the 48th Annual Meeting of the Association for Computational Linguistics, pages 678–687. find
  36. Stephen Tratz and Eduard Hovy. 2011. A Fast, Accurate, Non-Projective, Semantically-Enriched Parser. In Proceedings of the 2011 Conference on Empirical Methods in Natural Language Processing, pages 1257–1268. find
  37. Carl Vikner and Per Anker Jensen. 2002. A Semantic Analysis of the English Genitive. Interation of Lexical and Formal Semantics. Studia Linguistica, 56(2):191–226. find

Updated

↑ back to top

This page in Markdown → https://md.research.elamur.ai/free/en/papers/tratz-2013-automatic-interpretation-english-possessive/