Zero Pronoun Resolution in Japanese Discourse Based on Centering Theory

Summary

This paper tests centering theory on naturally occurring Japanese text for resolving zero pronouns (omitted obligatory arguments). Earlier centering work had been evaluated mostly on small sets of constructed discourses and covered only sequences of simple, single-predicate sentences, even though real text is full of complex sentences.

The authors split each complex sentence into its component simple sentences and run standard centering on the resulting sequence. This also lets intrasentential antecedents be handled. To make up for the shorter effective context after splitting, the search for antecedents is widened to the four preceding simple sentences, with closer clauses ranked higher. The method also uses the conjunctive postpositions linking adjacent clauses, which prevents counter-intuitive readings.

They implemented the approach on top of both Kameyama's and Walker's versions of centering and evaluated it on real Japanese discourses. The original algorithms reached roughly 60–70% accuracy, while the extended method reached about 78%. This suggests that surface-level, knowledge-light centering can scale to authentic text, although the evaluation is small.

Key points

Sentence complexity: a manual count found that 10,000 newspaper review-article sentences contain 24,332 simple sentences, so a real Japanese sentence averages about 2.0–2.5 clauses.
Antecedent distance: in one test discourse, 95% of antecedents were in the current or previous sentence. This matches Fujisawa et al.'s larger study, where 87.6% fell in the current or previous sentence and 95.1% within two preceding sentences.
Search range: accuracy rose as the window grew to about 2–4 previous simple sentences and then declined, so the authors fixed the window at four simple sentences.
Errors: in the Kameyama-based full system, 30.2% of wrong answers came from errors carried over from earlier resolutions. About 30 antecedents lay more than five sentences back, which the authors attribute to digressive sub-discourses (global discourse structure).
Comparison and scope: Takada and Doi's centering system scored 74.8% on zero pronouns. Zero pronouns whose antecedent never appears in the text (about 30%) were excluded from evaluation.

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: Discourse analysis · Coreference resolution

Cited by

12 papers of our library cite it (of the 140673 whose bibliographies we read); 5 of them have a page here, 7 not yet.

Cite

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

@inproceedings{okumura1996zero,
  title = {{Zero Pronoun Resolution in Japanese Discourse Based on Centering Theory}},
  author = {Okumura, Manabu and Tamura, Kouji},
  booktitle = {COLING 1996 Volume 2: The 16th International Conference on Computational Linguistics},
  year = {1996},
  doi = {10.3115/993268.993319},
  url = {https://doi.org/10.3115/993268.993319}
}

References

15 references, 1 of them with a card of ours. ↗ leads to the work, «find» searches for one we have no id for yet.

S.E. Brennan, M.W. Friedman, and C.J. Pollard, A Centering Approach to Pronouns, Proc. of the 25th Annual Meeting of the Association for Computational Linguistics, pp.155-162, 1987. find
S. l~Sjisawa, S. Masuymna, and S. Naito, A Basic Study on Ellipsis and Anaphora in Japanese Sentences, IPSJ SIG Notes, Vol. 91, No. 96, in Japanese, 1991. find
B.J. Grosz and C.L. Sidner, Attention, Intentions, and the Structure of Discourse, Computational Linguistics, Volume 12, Number 3, 175-204, 1986. find our card
B.J. Grosz, A.K. Joshi, and S. Weinstein, Centering: A Framework for Modeling tile Local Coherence of Discourse, Computational Linguistics, Volume 21, Nmuber 2, pp.203-225, 1995. find
J.R. Hobbs, Coherence and Coreference, Cogni-tire Science, Volume 3, Number 1, pp.67-90, 1979. find
M. Kameyama, A Property-Sharing Constraint in Centering, Proc. of the 2~th Annual Meeting of the Association for Computational Linguistics, pp.200-206, 1986. find
A. Kehler, The Effect of Establishing Coherence in Ellipsis and Anaphora Resolution, Proc. o] the 31st Annual Meeting of the Association for Computational Linguistics, pp.62-69, 1993. find
F. Minami, Gendai Nihongo no Kouzou, Taishuukan, Tokyo, in Japanese, 1974. find
H. Nakagawa and S. Nishizawa, Semantics of Complex Sentences in Japanese, Proe. of the 15th International Con]erence on Computational Linguistics, pp.679-685, 1994. find
H. Nakaiwa and S. Ikehara, Zero Pronoun Resolution in a Japanese to English Machine Translation System by using Verbal Semantic Attributes, Proe. of the 3rd Conference on Applied NatuTYtl Language PTveessing, pp.201-208, 1992. find
C.L. Sidner, Focusing in tile Comprehension of Definite Anaphora, In M. Brady and R.C. Berwick, editors, Computational Models of Discourse, MIT Press, pp.267-330, 1983. find
L.Z. Suri and K.F. McCoy, RAFT/RAPR and Centering: A Comparison and Discussion of Problems Related to Processing Complex Sentences, Computational Linguistics, Volume20, Number 2, pp.301-317, 1994. find
S. Takada and N. Doi, Centering in Japanese: A Step ]bwards Better Interpretation of Pronouns and Zero-Pronouns, Proe. of the 15th International Conference on Computational Linguistics, pp.1151-1156, 1994. find
M.A. Walker, M. Iida, and S. Cote, Japanese Discourse and the Process of Centering, Computational Linguistics, Volume 20, Number 2, pp.193-232, 1994. find
K. Yoshimoto, Study of Japanese Zero Pronouns in Discourse Processing, IPSJ SIG Notes, Vol. 86, No. 52, in Japanese, 1986. find

↑ back to top

This page in Markdown → https://md.research.elamur.ai/free/papers/okumura-1996-zero-pronoun-resolution-japanese-discourse/

📄 Zero Pronoun Resolution in Japanese Discourse Based on Centering Theory