Elamur Research⌕
☰
home / papers / 1996 / Bridging Textual Ellipses

Bridging Textual Ellipses

Summary

The paper addresses textual ellipsis (functional anaphora) in German. A definite noun phrase such as "Ladezeit" (charge time) refers back to an entity from the previous sentence (an accumulator) through an unstated conceptual relation, not through identity. A text understanding system must infer this link, or its knowledge base will be incoherent. The authors propose a hybrid resolver with two parts:

  • Conceptual criteria (language-independent): role chains are searched in a terminological domain knowledge base, and paths are ranked as plausible, metonymic or implausible.
  • Information-structure constraints (language-dependent): these come from a centering model adapted to German. Forward-looking centers are ordered by whether they are bound to the context, not by grammatical role.

The method is implemented in a dependency-based German text parser written in Smalltalk. On product-review texts from the information technology domain, it was correct in 95.2% of the cases it attempted. Recall was lower, mostly because the knowledge base lacked concepts. The work shows that encyclopedic domain knowledge combined with discourse structure can resolve bridging references.

Key points

  • Path well-formedness requires full connectivity, no cycles and no redundancy. Path patterns were derived from about 60 product reviews and from psycholinguistic findings. Metonymic patterns include producer-for-product, part-for-whole and whole-for-part.
  • Candidate antecedents are first compared by conceptual strength. When strength ties, the antecedent ranked higher in the Cf of the preceding utterance wins, with context-bound elements outranking unbound ones.
  • The evaluation knowledge base had 449 concepts and 334 relations. Of 109 ellipses, the system processed 63 and resolved 60 correctly: recall 55.05%, precision 95.2%.
  • Of 46 false negatives, 42 came from missing concept specifications and 4 from inappropriate triggering conditions.
  • The lexicon has about 3,000 entries and about 100 word classes, with concepts in LOOM (800 for IT, 500 for medicine). The parser was also tested on medical texts.

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

Topics: Coreference resolution

Cited by

9 papers of our library cite it (of the 140673 whose bibliographies we read); 4 of them have a page here, 5 not yet.

Cite

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

@inproceedings{hahn1996bridging,
  title = {{Bridging Textual Ellipses}},
  author = {Hahn, Udo and Strube, Michael and Markert, Katja},
  booktitle = {COLING 1996 Volume 1: The 16th International Conference on Computational Linguistics},
  year = {1996},
  doi = {10.3115/992628.992714},
  url = {https://doi.org/10.3115/992628.992714}
}

References

16 references, 0 of them with a card of ours. ↗ leads to the work, «find» searches for one we have no id for yet.

  1. Agha, G. & C. [lewitt (1987). Concurrent programming using actors. In A. Yonezawa & M. rlbkoro reds.), Object-Oriented Concurrent Progrwnming, pp. 37- 53. (?aml)ridge, Mass.: Mrl' Press. find
  2. Chaffin, R. (1992). The concept of semantic relations. In A. Imhrer & E. Kittay rEds.), Frames, Fields and Contrasts, pp. 253-288. Hillsdale, N.J.: 1,. Erlbaum. find
  3. Clark, H. H. (1975). Bridging. In TINLAP-I, pp. 169- 174. find
  4. Dane.~, F. (1974). Functkmal sentence perspective and tim organization of tim text. in F. Dancg rEd.), Papers on FunctionalSentence Perspective, pp. 11)6- 128. Prague: Academia. find
  5. Grosz, B. J., A. K. Joshi & S. Weinstein (1995). Center ing: A t}amework for modeling file local coherence of discourse. Computational Linguistics, 21 (2):203-225. find
  6. Hahr,, U. 11989). Making understanders out of parsers: Semantically driven parshlg as a key concept tot realistic text understanding applications. InternationalJoumral of Intelligent Systems, 4(3):345-393. Hahn, U., K. Markert & M. Strube 11996). A conceptual reasoning approach to the resolution of… find
  7. Hobbs, J. R., M. E. Stickel, D. E. Apl)elt & P. Mmlin (1993). Interpretation as abduction. Artificial lnteL ligence, 63:6%142. find
  8. Kamp, H. & U. Reyle (1993). [')'om Discourse to Logic. I)ordrecht: Kluwer. find
  9. Lakoft, G. (1987). Women, Nre, and Dangerous Things. Chicago: Chicago University Press. find
  10. MacGregor, R. & R. Bates (1987). The LOOM Knowledge RepresemationLanguage. (1S1/RS-87-18) USC/ISI. find
  11. Norvig, P. (1989). Marker passing as a weak melhod for inferencing. Cognitive Science, 13(4):569 620. find
  12. Pahner, M. S., D. A. Dahl, R. J. Schiffman & L. 1 lirschman (:1986). Recovering imt)licit infornlation, hi l'roc, of ACL-86, pp. 10-19. find
  13. Strube, M. & U. Hahn (1995). ParseTalk about sentence-and text-level atlaphora. In Prec. of EACL05, pp. 237-244. find
  14. Stnfl~e, M. & U. ltahn (:1996). Functional centering. In Prec. of ACL-96. find
  15. Wada, 1\|. (1994). A Irealment of functional dctinite descriptions. In Prec. ofCOLING-94, Vol. 2, pp. 789 -795. find
  16. Woods, W. A. & J. O. Schmolze (1992). The KL-ONE fmnily. Computers & Mathematics with Applications, 23(2-5): 133-177. find

↑ back to top

This page in Markdown →