Optimizing Referential Coherence in Text Generation
Paper · 2004 · Rodger Kibble and Richard Power
Summary
Kibble and Power address a weakness in natural language generation: generated texts often read poorly because clause order, argument order within clauses, and pronoun choice are decided separately. They argue that referential coherence should be a planning goal. Arranging facts and arguments well creates more chances for pronouns that readers can resolve without ambiguity. Using centering theory (CT), they implemented this idea in the Iconoclast document generator, which treats text structuring as constraint satisfaction. A user-supplied RST structure is expanded into candidate text structures that obey hard text-formation rules. These candidates are then ranked by weighted soft constraints. Among them are four centering preferences that recast CT's Rule 2: cohesion, salience, cheapness, and continuity. The article demonstrates the approach through sample outputs under one weighting scheme rather than through a quantitative evaluation. Its value is twofold. It offers a practical route to more fluent generated text in any RST-based generator with shallow lexicalization. It also serves as a research tool, because varying the constraint weights quickly shows how different CT variants affect the output.
Key points
- System and method: The approach is implemented in Iconoclast, which uses constraint satisfaction. Centering preferences are soft constraints weighted together with other stylistic preferences. They fix both the order of propositions and the order of arguments within clauses.
- Simplified CT settings:
- Entity ranking is reduced to subject > others.
- Only strict coreference counts as realization, so bridging references are excluded.
- Because of this, a "no-Cb" transition is allowed alongside Continue, Retain, Smooth Shift and Rough Shift.
- Preconditions for applying the method:
- Predicates map one-to-one onto verbs.
- Pronoun decisions are made only after grammatical roles and word order are settled.
- Position on related work: The authors favour weighting the centering principles over strictly ranking them, as proposed by Karamanis, Kibble 2001 and Beaver's OT account. They argue that strict ranking lacks empirical support.
- Evaluation reported elsewhere: Empirical tests appear in Kibble and Power 2004 (technical report RK/2004/1, ITRI-04-07). These comprise a paired-comparison study with naive judges, in which centering constraints noticeably improved acceptability, and a perturbation-based corpus study showing preferences among the constraints.
Institutions
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.Text generation · 202 works
Cite
Cite as Rodger Kibble and Richard Power. 2004. Optimizing Referential Coherence in Text Generation. Computational Linguistics.
@article{kibble2004optimizing,
title = {{Optimizing Referential Coherence in Text Generation}},
author = {Rodger Kibble and Richard Power},
journal = {Computational Linguistics},
year = {2004},
doi = {10.1162/0891201042544893},
url = {https://doi.org/10.1162/0891201042544893}
}References
- Beaver, David. 2004. The optimization of discourse anaphora. Linguistics and Philosophy, 27(1):3–56. find
- Brennan, Susan, Marilyn Walker Friedman, and Carl Pollard. 1987. A centering approach to pronouns. In Proceedings of 25th ACL, pages 155–162, Stanford, CA.
- Cahill, Lynne. 1999. Lexicalisation in applied NLG systems. Technical Report ITRI-99-04, Information Technology Research Institute, University of Brighton. find
- Callaway, Charles B. and James C. Lester. 2002. Pronominalization in generated discourse and dialogue. In Proceedings of the 40th Annual Meeting of the Association for Computational Linguistics (ACL), pages 88–95, Philadelphia. find
- Cheng, Hua. 2000. Experimenting with the interaction between aggregation and text planning. In Proceedings of ANLP-NAACL, pages 1–6, Seattle. find
- Cristea, Dan, Nancy Ide, and Laurent Romary. 1998. Veins theory: A model of global discourse cohesion and coherence. In Proceedings of COLING/ACL’98, pages 281–285, Montreal. find
- Grosz, Barbara, Aravind Joshi, and Scott Weinstein. 1995. Centering: A framework for modelling the local coherence of discourse. Computational Linguistics, 21(2):203–225. find
- Henschel, Renate, Hua Cheng, and Massimo Poesio. 2000. Pronominalisation revisited. In Proceedings of 18th COLING, pages 306–312, Saarbrücken, Germany. find
- Kameyama, Megumi. 1998. Intrasentential centering: A case study. In Marilyn Walker, Aravind Joshi, and Ellen Prince, editors, Centering Theory in Discourse, pages 89–112. Clarendon, Oxford. find
- Karamanis, Nikiforos. 2001. Exploring entity-based coherence. In Proceedings of Fourth CLUK, pages 18–26, University of Sheffield, Sheffield, England. find
- Kibble, Rodger. 1999. Cb or not Cb? Centering theory applied to NLG. In Proceedings of ACL Workshop on Discourse and Reference Structure, pages 72–81, University of Maryland, College Park. find
- Kibble, Rodger. 2001. A reformulation of rule 2 of centering theory. Computational Linguistics, 27(4):579–587.
- Kibble, Rodger. 2003. Towards the elimination of centering theory. In Ivana Kruijff-Korbayová and Claudia Kosny, editors, DiaBruck 2003: Proceedings of the Seventh Workshop on the Semantics and Pragmatics of Dialogue, Universität des Saarlandes, Saarbrücken, Germany. find
- Kibble, Rodger and Richard Power. 2004. Optimising referential coherence as a constraint satisfaction problem. Technical Report RK/2004/1, Department of Computing, Goldsmiths College, and ITRI-04-07, Information Technology Research Institute, University of Brighton. find
- Mann, William and Sandra Thompson. 1987. Rhetorical structure theory: A theory of text organisation. Technical Report ISI/RS-87-190, Information Sciences Institute, Los Angeles. find
- Marcu, Daniel. 1996. Building up rhetorical structure trees. In Proceedings of AAAI-96, pages 1069–1074, Portland, OR. find
- Marcu, Daniel. 1997. From local to global coherence: A bottom-up approach to text planning. In Proceedings of AAAI-97, pages 629–635, Providence, RI. find
- McCoy, Kathleen and Michael Strube. 1999. Generating anaphoric expressions: Pronoun or definite description? In Proceedings of ACL Workshop on Discourse and Reference Structure, pages 63–71, University of Maryland, College Park.
- McKeown, Kathleen R. 1985. Text Generation. Cambridge University Press, Cambridge. find
- Mellish, Chris, Alistair Knott, Jon Oberlander, and Mick O’Donnell. 1998. Experiments using stochastic search for text planning. In Proceedings of the Ninth International Workshop on Natural Language Generation, pages 97–108, Niagara-on-the-Lake, Ontario. find
- Mittal, Vibhu, Johanna Moore, Giuseppe Carenini, and Steven Roth. 1998. Describing complex charts in natural language: A caption generation system. Computational Linguistics, 24(3):431–467.
- Poesio, Massimo, Rosemary Stevenson, Hua Cheng, Barbara di Eugenio, and Janet Hitzeman. 2002. A corpus-based evaluation of centering theory. Technical Report TN-02-01/CSM-369, Natural Language Engineering Group, University of Essex. find
- Power, Richard. 2000. Planning texts by constraint satisfaction. In Proceedings of COLING 2000, pages 642–648, Saarbrücken, Germany. find
- Power, Richard, Donia Scott, and Nadjet Bouayad-Agha. 2003. Document structure. Computational Linguistics, 29(2):211–260. find
- Prince, Ellen. 1999. How not to mark topics: “Topicalization” in English and Yiddish. Unpublished manuscript, Linguistics Department, University of Pennsylvania. find
- Reiter, Ehud. 1994. Has a consensus NL generation architecture appeared, and is it psycholinguistically plausible? In Proceedings of the Seventh International Natural Language Generation Workshop, pages 163–170, Kennebunkport, ME. find
- Reiter, Ehud. 2000. Pipelines and size constraints. Computational Linguistics, 26(2):251–259. find
- Scott, Donia and Clarisse de Souza. 1990. Getting the message across in RST-based text generation. In Robert Dale, Chris Mellish, and Michael Zock, editors, Current Research in Natural Language Generation, pages 47–73. Academic Press, London.
- Strube, Michael and Udo Hahn. 1999. Functional centering—Grounding referential coherence in information structure. Computational Linguistics, 25(3):309–344.
- Suri, Linda, Kathleen McCoy, and Jonathan DeCristofaro. 1999. A methodology for extending focussing franeworks. Computational Linguistics, 25(2):173–194. van Hentenryck, P. 1989. Constraint Satisfaction in Logic Programming. MIT Press, Cambridge, MA. find
- Walker, Marilyn, Aravind Joshi, and Ellen Prince, editors. 1998a. Centering Theory in Discourse. Clarendon, Oxford. find
- Walker, Marilyn, Aravind Joshi, and Ellen Prince. 1998b. Centering in naturally occurring discourse. In Marilyn Walker, Aravind Joshi, and Ellen Prince, editors, Centering Theory in Discourse, pages 1–28. Clarendon, Oxford. find
Updated
This page in Markdown → https://md.research.elamur.ai/free/en/papers/kibble-2004-optimizing-referential-coherence-text-generation/