--- title: "Dynamic Coreference-Based Summarization — Baldwin 1998" description: "Baldwin and Morton address query-sensitive, indicative summarization: producing sentence extracts that let a user decide whether a retrieved document is…" authors: ["Breck Baldwin", "Thomas S. Morton"] year: 1998 venue: "Proceedings of the Third Conference on Empirical Methods for Natural Language Processing" source: "https://aclanthology.org/W98-1501/" licence: "CC BY-NC-SA 3.0 (ACL Anthology, before 2016)" licence_url: "https://creativecommons.org/licenses/by-nc-sa/3.0/" summary_licence: "Elamur Cosmic License" canonical: "https://research.elamur.ai/free/papers/baldwin-1998-dynamic-coreference-based-summarization/" --- [Elamur Research](https://md.research.elamur.ai/) · [this page in HTML](https://research.elamur.ai/free/papers/baldwin-1998-dynamic-coreference-based-summarization/) [home](https://md.research.elamur.ai/) / [papers](https://md.research.elamur.ai/free/papers/) / [1998](https://md.research.elamur.ai/free/papers/1998/) / Dynamic Coreference-Based Summarization # Dynamic Coreference-Based Summarization **Authors:** [Breck Baldwin](https://md.research.elamur.ai/free/authors/breck-baldwin/), [Thomas S. Morton](https://md.research.elamur.ai/free/authors/thomas-s-morton/) · **Year:** 1998 · **Venue:** Proceedings of the Third Conference on Empirical Methods for Natural Language Processing · **Licence:** [CC BY-NC-SA 3.0 (ACL Anthology, before 2016)](https://creativecommons.org/licenses/by-nc-sa/3.0/) **Links:** [Publisher's page](https://aclanthology.org/W98-1501/) · [PDF](https://aclanthology.org/W98-1501.pdf) · [OpenAlex](https://openalex.org/W63003940) **Citations:** 64 citations (openalex-snapshot · 2026-10-14) · [9](#cited-by) in our library, of the 140673 papers whose bibliographies we read ## Summary Baldwin and Morton address query-sensitive, indicative summarization: producing sentence extracts that let a user decide whether a retrieved document is relevant without reading all of it. Their system is built on the CAMP NLP system. It links phrases in the query and headline to phrases in the document through a limited set of coreference-style relations: - identity between entities, including acronyms, - identity between events, expressed as verbs or nominalizations, - part–whole relations, such as a city or province belonging to a country named in the query. Sentences are ranked by how many uncovered query chains they contain, then by headline chains, then by other chains. Headline chains are included so the extract does not misrepresent the document. Sentences are added until every query chain is covered. Relevance judgments made from these summaries came close to judgments made from full documents, while the summaries kept only about 17% of the text. The result suggests that shallow coreference reasoning offers a practical middle ground between plain string matching and deeper understanding. ## Key points - A preliminary analysis of the TIPSTER dry run (5 queries × 200 documents, 6 systems) found that the union of the systems' outputs gave a correct summary for nearly every relevant document. This suggested the task is feasible with existing technology. - Two verbs are associated mainly through mutual information, with a threshold of 5. The statistics come from about 45,000 Wall Street Journal articles. Very frequent verbs are filtered out unless they share a subject or object with the query verb. - Sentences are selected with a seven-level tie-breaking ranking. Selection stops once all query chains are covered and the summary has at least 4 sentences. - First evaluation: precision was about 82.8% (101/122) and compression about 82.8%. Agreement with the TREC assessors was reported as roughly 76–79%, against an estimated 80.1% agreement between two human assessors. - Second evaluation (910 documents): precision 80.3%, recall 57.6%, accuracy 65.3%, compression 83.0%. For comparison, the same assessors judging full documents reached 83.5% precision and about 69% accuracy. *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](https://md.research.elamur.ai/free/shelves/linguistics/) Topics: [Coreference resolution](https://md.research.elamur.ai/free/topics/coreference-resolution/) ## Cited by 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. - [Using Coreference Chains for Text Summarization](https://md.research.elamur.ai/free/papers/azzam-1999-coreference-chains-text-summarization/) - [Gesture Improves Coreference Resolution](https://md.research.elamur.ai/free/papers/eisenstein-2006-gesture-improves-coreference-resolution/) - [Using Coreference for Question Answering](https://md.research.elamur.ai/free/papers/morton-1999-coreference-question-answering/) - [A Probabilistic Annotation Model for Crowdsourcing Coreference](https://md.research.elamur.ai/free/papers/paun-2018-probabilistic-annotation-model-crowdsourcing-coreference/) ## Cite The paper itself, not this page. BibTeX from the library's record of the work. ```bibtex @inproceedings{baldwin1998dynamic, title = {{Dynamic Coreference-Based Summarization}}, author = {Baldwin, Breck and Morton, Thomas S.}, booktitle = {Proceedings of the Third Conference on Empirical Methods for Natural Language Processing}, year = {1998}, url = {https://aclanthology.org/W98-1501/} } ``` - [← previous DiMLex: A Lexicon of Discourse Markers for Text Generation and Understanding](https://md.research.elamur.ai/free/papers/stede-1998-dimlex-lexicon-discourse-markers-text/) - [next → Evaluating a Focus-Based Approach to Anaphora Resolution](https://md.research.elamur.ai/free/papers/azzam-1998-evaluating-focus-based-approach-anaphora/) - [↑ back to papers of 1998](https://md.research.elamur.ai/free/papers/1998/) ## References 11 references, 0 of them with a card of ours. ↗ leads to the work, «find» searches for one we have no id for yet. 1. [I] Breck Baldwin. CogNIAC: High precision coref-enmce with limited knowledge and linguistie resources. In Pmccedings of the ACL WoTkshop on Openthono..l F'acf;ors in Pnzct?:cal, Rolmst Anaphom n:solni,ion for UnTesf,Tic.ted Texts) pages 38---45) iVJ.a.drid) Spain) .June 1997. [find](https://scholar.google.com/scholar?q=%5BI%5D+Breck+Baldwin%2E+CogNIAC%3A+High+precision+coref%2Denmce+with+limited+knowledge+and+linguistie+resources%2E+In+Pmccedings+of+the+ACL+WoTkshop+on+Openthono%2E%2El+F%27acf%3Bors+in+Pnzct%3F%3Acal%2C+Rolmst+Anaphom+n%3Asoln) 2. [2] Breck Baldwin, Christine Doran, .Jeffrey C. Rey-ua.r) I\\!Jicha.el Niv) B. Srinivas) and iVIark Wasson. EAGLE: An extensible architecture for general linguistic engineering. In Proceedings of RIA0~97, lVfontreal) 1997. [find](https://scholar.google.com/scholar?q=%5B2%5D+Breck+Baldwin%2C+Christine+Doran%2C+%2EJeffrey+C%2E+Rey%2Dua%2Er%29+I%5C%5C%21Jicha%2Eel+Niv%29+B%2E+Srinivas%29+and+iVIark+Wasson%2E+EAGLE%3A+An+extensible+architecture+for+general+linguistic+engineering%2E+In+Proceedings+of+RIA0) 3. [3] Michael Bett and .J adc Goldstein. Automated query-relevant document summarization. In Pm-ceedings of Tipster· Te:Dt Phase III 12-Month Workshop, 1997. [find](https://scholar.google.com/scholar?q=%5B3%5D+Michael+Bett+and+%2EJ+adc+Goldstein%2E+Automated+query%2Drelevant+document+summarization%2E+In+Pm%2Dceedings+of+Tipster%C2%B7+Te%3ADt+Phase+III+12%2DMonth+Workshop%2C+1997%2E) 4. [4] Michael Chrzanowski, Therese Firmin, Lynette Hirsclunan) David House, Inderjec\\L Nia.ni) Leo Obrst) Sara Shelton) Beth Sundheim 1 and Sandra Wagner. (SUMMAC) call for participation. http:/ j www. tipster.org/ summca.ll.htm, Jan-nary 1998. [find](https://scholar.google.com/scholar?q=%5B4%5D+Michael+Chrzanowski%2C+Therese+Firmin%2C+Lynette+Hirsclunan%29+David+House%2C+Inderjec%5C%5CL+Nia%2Eni%29+Leo+Obrst%29+Sara+Shelton%29+Beth+Sundheim+1+and+Sandra+Wagner%2E+%28SUMMAC%29+call+for+participation%2E+http%3A%2F+j+www%2E) 5. [5] Michael John Collins. A New Statistical Parser Based on Digram Lexical Dependencies. In Proceedings of the 34th Anmwl Meeting of the ACL, 1996 [find](https://scholar.google.com/scholar?q=%5B5%5D+Michael+John+Collins%2E+A+New+Statistical+Parser+Based+on+Digram+Lexical+Dependencies%2E+In+Proceedings+of+the+34th+Anmwl+Meeting+of+the+ACL%2C+1996) 6. [6] Therese Hand. Tipster summarization evaluation taslcdry-run evaluation results. In Proceedings of TipsteT Te:ct Phose III 12-Month Wor-kshop, !997 . [7] Daniel Karp) Yves Schabes Martin Zaidel, and Dania Egedi. A freely available wide eovcrage morphological analyzer for english. In P1-occedings… [find](https://scholar.google.com/scholar?q=%5B6%5D+Therese+Hand%2E+Tipster+summarization+evaluation+taslcdry%2Drun+evaluation+results%2E+In+Proceedings+of+TipsteT+Te%3Act+Phose+III+12%2DMonth+Wor%2Dkshop%2C+%21997+%2E+%5B7%5D+Daniel+Karp%29+Yves+Schabes+Martin+Zaidel%2C+an) 7. [8] Ad wait Ratnaparkhi. A Maximum Entropy Part of Speech Tagger. In Eric Brill cul.d Kenneth Church) editors) Confen:nce on J-i}mpin:co,l MdJwrls -in Nat-nml Langnage Process1\>n,_r]) University of Pennsylvania1 May 1.7~18 1996. [find](https://scholar.google.com/scholar?q=%5B8%5D+Ad+wait+Ratnaparkhi%2E+A+Maximum+Entropy+Part+of+Speech+Tagger%2E+In+Eric+Brill+cul%2Ed+Kenneth+Church%29+editors%29+Confen%3Ance+on+J%2Di%7Dmpin%3Aco%2Cl+MdJwrls+%2Din+Nat%2Dnml+Langnage+Process1%5C%3En%2C%5Fr%5D%29+University+of+P) 8. [9] Jdfrcy C. Il.\\<\>ynilr and Adwait Ratna.parkhi. A maximum entropy approach to identifying sen~ tence boundaries. In Proceedings of the Fifth Conference on Applied NoJnntl Lang'ltage Processing, pages 16 19, Washington, D.C., April 1997. [find](https://scholar.google.com/scholar?q=%5B9%5D+Jdfrcy+C%2E+Il%2E%5C%3C%5C%3Eynilr+and+Adwait+Ratna%2Eparkhi%2E+A+maximum+entropy+approach+to+identifying+sen%7E+tence+boundaries%2E+In+Proceedings+of+the+Fifth+Conference+on+Applied+NoJnntl+Lang%27ltage+Processing%2C+pa) 9. [10] Tomek Strzalkowski) F~~ng Lin) .Jin \\Nang) Langdon VVhitc) and Bowden \\~lise. Natural language information retrieval and summarization. In Proceedings of Tipster Te.?:i Phase III 12-Month Workshop, 1997. [find](https://scholar.google.com/scholar?q=%5B10%5D+Tomek+Strzalkowski%29+F%7E%7Eng+Lin%29+%2EJin+%5C%5CNang%29+Langdon+VVhitc%29+and+Bowden+%5C%5C%7Elise%2E+Natural+language+information+retrieval+and+summarization%2E+In+Proceedings+of+Tipster+Te%2E%3F%3Ai+Phase+III+12%2DMonth+Works) 10. [11 J Ellen M. Voorhees and Donna Hannan. Overview of the fifth Text REtrieval Conference (TREC-5). In Pmccedings of the Fifth Te.xl; REtr·icval Conference {TREC-5}, pages I 28. NIST 500-238, 1997. [find](https://scholar.google.com/scholar?q=%5B11+J+Ellen+M%2E+Voorhees+and+Donna+Hannan%2E+Overview+of+the+fifth+Text+REtrieval+Conference+%28TREC%2D5%29%2E+In+Pmccedings+of+the+Fifth+Te%2Exl%3B+REtr%C2%B7icval+Conference+%7BTREC%2D5%7D%2C+pages+I+28%2E+NIST+500%2D238%2C+1997%2E) 11. [12] Nina Wacholder, Yael R.avin and Misook Choi. Disambiguation of proper mt.mes in text. In Pro-cecdi,ngs of t.he Fifth Confenmcc o·n Applied Natu-ro,l Lan,rrnage Pmccssi11.g 1 l\\tlay 1997. [find](https://scholar.google.com/scholar?q=%5B12%5D+Nina+Wacholder%2C+Yael+R%2Eavin+and+Misook+Choi%2E+Disambiguation+of+proper+mt%2Emes+in+text%2E+In+Pro%2Dcecdi%2Cngs+of+t%2Ehe+Fifth+Confenmcc+o%C2%B7n+Applied+Natu%2Dro%2Cl+Lan%2Crrnage+Pmccssi11%2Eg+1+l%5C%5Ctlay+1997%2E)