--- title: "Automatic Pronominal Anaphora Resolution in English Texts (2004)" description: "The paper addresses pronominal anaphora resolution in English: linking pronouns such as he , her , itself or… 89% accuracy. 17 references." authors: ["Tyne Liang", "Dian-Song Wu"] year: 2004 venue: "International Journal of Computational Linguistics & Chinese Language Processing, Volume 9, Number 1, February 2004: Special Issue on Selected Papers from ROCLING XV" source: "https://aclanthology.org/O04-2002/" 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/en/papers/liang-2004-automatic-pronominal-anaphora-resolution-english/" --- [Elamur Research](https://md.research.elamur.ai/free/en/) · [this page in HTML](https://research.elamur.ai/free/en/papers/liang-2004-automatic-pronominal-anaphora-resolution-english/) [home](https://md.research.elamur.ai/free/en/) / [papers](https://md.research.elamur.ai/free/en/papers/) / [2004](https://md.research.elamur.ai/free/en/papers/2004/page/2/) / Automatic Pronominal Anaphora Resolution in English Texts # Automatic Pronominal Anaphora Resolution in English Texts **Authors:** [Tyne Liang](https://md.research.elamur.ai/free/en/authors/tyne-liang/) and [Dian-Song Wu](https://md.research.elamur.ai/free/en/authors/dian-song-wu/) · **Year:** 2004 · **Venue:** [International Journal of Computational Linguistics & Chinese Language Processing, Volume 9, Number 1, February 2004: Special Issue on Selected Papers from ROCLING XV](https://md.research.elamur.ai/free/en/venues/rocling/) ([ROCLING 2004](https://md.research.elamur.ai/free/en/venues/rocling/2004/)) · **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/O04-2002/) · [PDF](https://aclanthology.org/O04-2002.pdf) · [OpenAlex](https://openalex.org/W2138529602) **Citations:** 27 citations (OpenAlex, 2026-10-14) ## Summary The paper addresses pronominal anaphora resolution in English: linking pronouns such as *he*, *her*, *itself* or *their* to an earlier noun phrase, both inside a sentence and across sentences. The pipeline is knowledge-light and needs no full parsing: - texts are POS-tagged with the TOSCA-ICLE tagger; - base noun phrases are found by a finite-state machine; - non-referring ("pleonastic") *it* is filtered out first. Candidates within a window are then pruned by number, gender and animacy agreement. Gender comes from a name gazetteer plus WordNet. Animacy is decided by tracing a noun's hypernyms up to its WordNet unique beginner, with the governing verb as an extra hint. Survivors are ranked by seven heuristic rules (five preference, two constraint), and any agreement failure zeroes the score. On Brown corpus texts from several genres, the system picks the correct antecedent 77% of the time, above the 50–62% reported for earlier knowledge-poor systems. This shows that cheap semantic animacy cues can rival approaches that rely on full syntactic parses. ## Key points - Overall success rate is 77%. By genre: Reportage and Editorial 80%, Reviews and Fiction 79%, Skills 78%, Religion 76%, Lore 69%. The paper's comparison table lists Kennedy and Boguraev at 62%, Mitkov at 57% and Baldwin's CogNIAC at 50%. - Pleonastic *it* makes up about 9% of pronouns, and the dedicated module detects it with 89% accuracy. Detection relies on weather/time patterns, modal adjectives (including comparative and superlative forms) and cognitive verbs. - In the Brown corpus, roughly 60% of anaphors have their antecedent in the same sentence. This differs from corpora containing only *it*/*its*, used in earlier evaluations. - The heuristic rule layer removes about 30% of the remaining errors. The animacy module helps most in the reportage genre. - POS tagging is about 95% accurate. A noun mistagged as another word class (e.g. *patient* as an adjective) can never become a candidate. Coordinated multi-entity antecedents and unknown words are the other main error sources. ## Interesting ideas - Animacy is decided by walking a noun's hypernym chain to its WordNet unique beginner, then cross-checking with the verb the entity governs, since some verbs strongly imply a living agent. - The scoring formula multiplies the rule score by hard agreement factors, so a single number, gender or animacy mismatch drives the candidate to zero, however many preferences it satisfies. - Expletive *it* is detected and removed before resolution even starts, using small lexical pattern lists (weather/time, modal adjectives, cognitive verbs). - Noun phrases are split into minimal base NPs for candidate selection, then the full modifier is reassembled once the antecedent is chosen. This keeps prepositional attachments from hiding the true head. - Evaluation is broken down by genre and by the error rate after each successive filter (number, gender, animacy, heuristics), which shows each module's contribution. ## For Mova How these ideas could serve Mova, our small open language model — its files, stages and tests: - **I1** (frames): Our coreference step could use a WordNet-root animacy feature plus a verb-based check from VerbNet selectional restrictions to filter pronoun candidates in a fully explainable way. - **I2** (frames): Coreference ranking could keep constraints as multiplicative vetoes and preferences as additive points, so every rejected antecedent has a named, traceable reason. - **I3** (world): Marking expletive *it* early would stop phantom entities from being created in the world of the text and avoid spurious holdings or states attached to them. - **I4** (world): When creating entities from UD subtrees, we could resolve against the head-level NP and only afterwards attach the full span as the entity's description. - **I5** (learn): Coreference evaluation could report per-stage ablation curves on fables, school texts and other genres, revealing which filter actually earns its keep. *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 - [National Yang Ming Chiao Tung University](https://md.research.elamur.ai/free/en/universities/national-yang-ming-chiao-tung-university/) — University, Hsinchu, Taiwan · 5 works · 85 citations 📚 Shelves (4): [🖥️ Computer Science](https://md.research.elamur.ai/free/en/shelves/computer-science/) · 16,667 works · [🤖 Artificial Intelligence](https://md.research.elamur.ai/free/en/shelves/computer-science/artificial-intelligence/) · 16,194 works · [🎭 Arts and Humanities](https://md.research.elamur.ai/free/en/shelves/arts-and-humanities/) · 16,039 works · [🗣️ Language and Linguistics](https://md.research.elamur.ai/free/en/shelves/arts-and-humanities/language-and-linguistics/) · 15,891 works 🧭 Topics (2): [English](https://md.research.elamur.ai/free/en/topics/english/) · 2,263 works · [🔗 Coreference resolution](https://md.research.elamur.ai/free/en/topics/coreference-resolution/) · 574 works ## Cite Cite as: Tyne Liang and Dian-Song Wu. 2004. Automatic Pronominal Anaphora Resolution in English Texts. International Journal of Computational Linguistics & Chinese Language Processing, Volume 9, Number 1, February 2004: Special Issue on Selected Papers from ROCLING XV. The paper itself, not this page. BibTeX from the library's record of the work. ```bibtex @article{liang2004automatic, title = {{Automatic Pronominal Anaphora Resolution in English Texts}}, author = {Tyne Liang and Dian-Song Wu}, journal = {International Journal of Computational Linguistics \& Chinese Language Processing, Volume 9, Number 1, February 2004: Special Issue on Selected Papers from ROCLING XV}, year = {2004}, doi = {10.30019/ijclclp.200402.0002}, url = {https://doi.org/10.30019/ijclclp.200402.0002} } ``` - [← previous A High-Performance Coreference Resolution System using a Constraint-based Multi-Agent Strategy](https://md.research.elamur.ai/free/en/papers/zhou-2004-high-performance-coreference-resolution-system/) - [next → Synchronous Grammars as Tree Transducers](https://md.research.elamur.ai/free/en/papers/shieber-2004-synchronous-grammars-as-tree-transducers/) - [↑ back to papers of 2004](https://md.research.elamur.ai/free/en/papers/2004/page/2/) ## References 17 references, 9 of them with a card of ours. ↗ leads to the work, «find» searches for one we have no id for yet. 1. Jan, Aarts, Henk Barkema and Nelleke Oostdijk, “The TOSCA-ICLE Tagset: Tagging Manual,” TOSCA Research Group for Corpus Linguistics, 1997. [find](https://scholar.google.com/scholar?q=Jan%2C+Aarts%2C+Henk+Barkema+and+Nelleke+Oostdijk%2C+%E2%80%9CThe+TOSCA%2DICLE+Tagset%3A+Tagging+Manual%2C%E2%80%9D+TOSCA+Research+Group+for+Corpus+Linguistics%2C+1997%2E) 2. Baldwin, Breck, “[CogNIAC: high precision coreference with limited knowledge and linguistic resources](https://md.research.elamur.ai/free/en/papers/baldwin-1997-cogniac-high-precision-coreference-limited/),” In Proceedings of the ACL'97/EACL'97 workshop on Operational factors in practical, robust anaphora resolution, 1997, pp. 38-45. 3. Tyne Liang and Dian-Song Wu [Automatic Pronominal Anaphora Resolution in English Texts](https://md.research.elamur.ai/free/en/papers/liang-2003-automatic-pronominal-anaphora-resolution-english/) Bontcheva, Kalina, Marin Dimitrov, Diana Maynard and Valentin Tablan, “Shallow Methods for Named Entity Coreference Resolution,” In Proceedings of TRAITEMENT AUTOMATIQUE DES LANGUES NATURELLES (TALN), 2002, pp. 24-32. 4. Cardie, Claire and Kiri Wagstaff, “[Noun Phrase Coreference as Clustering](https://md.research.elamur.ai/free/en/papers/cardie-1999-noun-phrase-coreference-as-clustering/),” In Proceedings of the Joint Conference on Empirical Methods in Natural Language Processing and Very Large Corpora, 1999. 5. Chen, Kuang-hua and Hsin-Hsi Chen, “Extracting Noun Phrases from Large-Scale Texts: A Hybrid Approach and Its Automatic Evaluation,” In Proceedings of the 32nd ACL Annual Meeting, 1994, pp. 234-241. [find](https://scholar.google.com/scholar?q=Chen%2C+Kuang%2Dhua+and+Hsin%2DHsi+Chen%2C+%E2%80%9CExtracting+Noun+Phrases+from+Large%2DScale+Texts%3A+A+Hybrid+Approach+and+Its+Automatic+Evaluation%2C%E2%80%9D+In+Proceedings+of+the+32nd+ACL+Annual+Meeting%2C+1994%2C+pp%2E+234%2D241%2E) 6. Dagan, Ido and Alon Itai, “[Automatic processing of large corpora for the resolution of anaphora references](https://md.research.elamur.ai/free/en/papers/dagan-1990-automatic-processing-large-corpora-resolution/),” In Proceedings of the 13th International Conference on Computational Linguistics (COLING'90), Vol. III, 1-3, 1990. 7. Denber, Michel, “Automatic resolution of anaphora in English,” Technical report, Eastman Kodak Co. , 1998. [find](https://scholar.google.com/scholar?q=Denber%2C+Michel%2C+%E2%80%9CAutomatic+resolution+of+anaphora+in+English%2C%E2%80%9D+Technical+report%2C+Eastman+Kodak+Co%2E+%2C+1998%2E) 8. Evans, Richard and Constantin Orasan, “Improving anaphora resolution by identifying animate entities in texts,” In Proceedings of DAARC, 2000. [find](https://scholar.google.com/scholar?q=Evans%2C+Richard+and+Constantin+Orasan%2C+%E2%80%9CImproving+anaphora+resolution+by+identifying+animate+entities+in+texts%2C%E2%80%9D+In+Proceedings+of+DAARC%2C+2000%2E) 9. Ge, Niyu, John Hale and Eugene Charniak, “[A Statistical Approach to Anaphora Resolution](https://md.research.elamur.ai/free/en/papers/ge-1998-statistical-approach-anaphora-resolution/),” In Proceedings of the Sixth Workshop on Very Large Corpora (COLING-ACL98), 1998, pp.161-170. 10. Kennedy, Christopher and Branimir Boguraev, “[Anaphora for everyone: Pronominal anaphora resolution without a parser](https://md.research.elamur.ai/free/en/papers/kennedy-1996-anaphora-everyone-pronominal-anaphora-resolution/),” In Proceedings of the 16 th International Conference on Computational Linguistics, 1996, pp.113-118. 11. Lappin, Shalom and Herbert Leass, “[An Algorithm for Pronominal Anaphora Resolution](https://md.research.elamur.ai/free/en/papers/lappin-1994-algorithm-pronominal-anaphora-resolution/),” Computational Linguistics, Volume 20, Part 4, 1994, pp. 535-561. [↗](https://doi.org/10.5555/203987.203989) 12. Miller, George, “Nouns in WordNet: A Lexical Inheritance System,” Journal of Lexicography, 1993, pp. 245-264. [find](https://scholar.google.com/scholar?q=Miller%2C+George%2C+%E2%80%9CNouns+in+WordNet%3A+A+Lexical+Inheritance+System%2C%E2%80%9D+Journal+of+Lexicography%2C+1993%2C+pp%2E+245%2D264%2E) 13. Mitkov, Ruslan, “[Robust pronoun resolution with limited knowledge](https://md.research.elamur.ai/free/en/papers/mitkov-1998-robust-pronoun-resolution-limited-knowledge/), “ In Proceedings of the 18th International Conference on Computational Linguistics (COLING'98)/ACL'98 Conference Montreal, Canada. 1998, pp. 869-875. 14. Mitkov, Ruslan, “Anaphora Resolution: The State of the Art,” Working paper (Based on the COLING'98/ACL'98 tutorial on anaphora resolution), 1999. [find](https://scholar.google.com/scholar?q=Mitkov%2C+Ruslan%2C+%E2%80%9CAnaphora+Resolution%3A+The+State+of+the+Art%2C%E2%80%9D+Working+paper+%28Based+on+the+COLING%2798%2FACL%2798+tutorial+on+anaphora+resolution%29%2C+1999%2E) 15. Mitkov, Ruslan and Catalina Barbu, “[Evaluation tool for rule-based anaphora resolution methods](https://md.research.elamur.ai/free/en/papers/barbu-2001-evaluation-tool-rule-based-anaphora/),” In Proeedings of ACL'01, Toulouse, 2001. 16. Mitkov, Ruslan, Richard Evans and Constantin Orasan, “A new fully automatic version of Mitkov's knowledge-poor pronoun resolution method,” In Proceedings of CICLing-2000, Mexico City, Mexico. [find](https://scholar.google.com/scholar?q=Mitkov%2C+Ruslan%2C+Richard+Evans+and+Constantin+Orasan%2C+%E2%80%9CA+new+fully+automatic+version+of+Mitkov%27s+knowledge%2Dpoor+pronoun+resolution+method%2C%E2%80%9D+In+Proceedings+of+CICLing%2D2000%2C+Mexico+City%2C+Mexico%2E) 17. Wang, Ning Yuan, Chunfa, Wang, K.F. and Li, Wenjie “Anaphora Resolution in Chinese Financial News for Information Extraction,” In Proceedings of 4th World Congress on Intelligent Control and Automation, June 2002, pp.2422-2426. [find](https://scholar.google.com/scholar?q=Wang%2C+Ning+Yuan%2C+Chunfa%2C+Wang%2C+K%2EF%2E+and+Li%2C+Wenjie+%E2%80%9CAnaphora+Resolution+in+Chinese+Financial+News+for+Information+Extraction%2C%E2%80%9D+In+Proceedings+of+4th+World+Congress+on+Intelligent+Control+and+Automatio) *Updated 1 November 2026*