---
title: "Exploiting Syntactico-Semantic Structures for Relation Extraction"
description: "The paper argues that relation extraction (RE) has a second axis that is independent of the… 3.6 points. 14 references, cited by 73 papers here."
authors: ["Yee Seng Chan", "Dan Roth"]
year: 2011
venue: "Proceedings of the 49th Annual Meeting of the Association for Computational Linguistics: Human Language Technologies"
source: "https://aclanthology.org/P11-1056/"
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/chan-2011-exploiting-syntactico-semantic-structures-relation/"
---
[Elamur Research](https://md.research.elamur.ai/free/en/) · [this page in HTML](https://research.elamur.ai/free/en/papers/chan-2011-exploiting-syntactico-semantic-structures-relation/)

[home](https://md.research.elamur.ai/free/en/) / [papers](https://md.research.elamur.ai/free/en/papers/) / [2011](https://md.research.elamur.ai/free/en/papers/2011/) / Exploiting Syntactico-Semantic Structures for Relation Extraction

# Exploiting Syntactico-Semantic Structures for Relation Extraction

**Authors:** [Yee Seng Chan](https://md.research.elamur.ai/free/en/authors/yee-seng-chan/) and [Dan Roth](https://md.research.elamur.ai/free/en/authors/dan-roth/) · **Year:** 2011 · **Venue:** [Proceedings of the 49th Annual Meeting of the Association for Computational Linguistics: Human Language Technologies](https://md.research.elamur.ai/free/en/venues/acl/) ([ACL 2011](https://md.research.elamur.ai/free/en/venues/acl/2011/)) · **Publisher:** [Association for Computational Linguistics](https://md.research.elamur.ai/free/en/publishers/association-for-computational-linguistics/) · **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/P11-1056/) · [PDF](https://aclanthology.org/P11-1056.pdf) · [OpenAlex](https://openalex.org/W2132516856)

**Citations:** 179 citations (OpenAlex, 2026-10-14) · [73](https://md.research.elamur.ai/free/en/papers/chan-2011-exploiting-syntactico-semantic-structures-relation/cited-by/) in our library, of the 141,840 papers whose bibliographies we read

## Summary
The paper argues that relation extraction (RE) has a second axis that is independent of the relation's semantic type: the syntactico-semantic form in which a relation is expressed. Using the ACE-2004 corpus, where annotators recorded this form, the authors take four structures (premodifier, possessive, preposition, formulaic) and detect them with simple, mostly syntactic rules over POS tags, mention boundaries, dependency labels and coarse entity types. Mention pairs that match a structure go to a dedicated classifier, RE s, which is trained only on such pairs. All other pairs go to a standard classifier, RE base. When mentions must be predicted rather than given, the structures also help rescue candidate mentions that the entity typing model rejected with low confidence, which limits error propagation in the pipeline. The approach gives consistent gains, including F1 improvements of 8.2, 4.6 and 3.6 points on predicted mentions. Gains are especially notable when training data is scarce.

## Key points
- Data: ACE-2004 with 345 documents and cross-validation folds of 276 training documents each. Of 55,520 within-sentence mention pairs, only 4,011 (about 7.2%) hold a valid relation.
- Structure distribution among the related pairs is: premodifier 30.5%, possessive 18.9%, preposition 23.9%, formulaic 7.2% and verbal 19.5%. The four structures handled cover about 80% of relations.
- Pattern recall and precision (%) are: PreMod 86.8/79.7, Poss 94.3/88.3, Prep 94.6/20.0 and Formula 85.5/62.2. Prepositions are hardest because of attachment ambiguity, preposition sense ambiguity and multiple intervening prepositions.
- A fine-grained mention entity typing (MET) classifier with 44 labels (43 types plus null) is trained with LIBLINEAR. A candidate rejected as a non-mention is kept if its null probability is at most t=0.2; the authors later found that t=0.25 worked better.
- With predicted mentions, adding the patterns raised F1 by 8.2 (binary), 4.6 (coarse) and 3.6 (fine-grained). Improvements held across all training-set sizes, from 5% to 80% of the data.

## Interesting ideas
- **Factor the task along an orthogonal axis.** The authors first classify how a relation is expressed and only then what type it is. This kind of split could make other sparse classification tasks easier, because cheap structural filters remove most negatives and leave a more balanced training set.
- **Specialized model plus fallback.** One classifier is trained only on structurally matched pairs, and a general model handles everything else. This simple routing scheme could suit any task where one subpopulation is easy to recognize and behaves differently from the rest.
- **Recovering low-confidence pipeline rejections.** Candidates that an upstream component rejected with low confidence are reconsidered when downstream evidence (here, a structural pattern) supports them. This is a lightweight form of joint inference that reduces cascading errors without retraining the pipeline.
- **Exploit annotator metadata.** ACE-2004 marked each relation's structure type, which allowed the patterns to be measured for recall and precision directly. Checking existing treebanks or corpora for such side annotations can provide free supervision or evaluation signals.
- **Hand-written patterns as a low-data aid.** Because the structures are tightly constrained, a few rules give substantial benefit when labelled data is limited. This is relevant for lower-resource languages, where parsers and POS taggers exist but annotated relation data is scarce.

*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

- [University of Illinois Urbana-Champaign](https://md.research.elamur.ai/free/en/universities/university-of-illinois-urbana-champaign/) — University, Urbana, United States · 108 works · 67,432 citations

📚 Shelves (4): [🖥️ Computer Science](https://md.research.elamur.ai/free/en/shelves/computer-science/) · 9,085 works · [🤖 Artificial Intelligence](https://md.research.elamur.ai/free/en/shelves/computer-science/artificial-intelligence/) · 8,888 works · [🎭 Arts and Humanities](https://md.research.elamur.ai/free/en/shelves/arts-and-humanities/) · 9,296 works · [🗣️ Language and Linguistics](https://md.research.elamur.ai/free/en/shelves/arts-and-humanities/language-and-linguistics/) · 9,206 works

🧭 Topics (1): [🔎 Information extraction](https://md.research.elamur.ai/free/en/topics/information-extraction/) · 109 works

## Cite

Cite as: Yee Seng Chan and Dan Roth. 2011. Exploiting Syntactico-Semantic Structures for Relation Extraction. Proceedings of the 49th Annual Meeting of the Association for Computational Linguistics: Human Language Technologies.

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

```bibtex
@inproceedings{chan2011exploiting,
  title = {{Exploiting Syntactico-Semantic Structures for Relation Extraction}},
  author = {Yee Seng Chan and Dan Roth},
  booktitle = {Proceedings of the 49th Annual Meeting of the Association for Computational Linguistics: Human Language Technologies},
  year = {2011},
  url = {https://aclanthology.org/P11-1056/}
}
```

- [← previous Learning Discriminative Projections for Text Similarity Measures](https://md.research.elamur.ai/free/en/papers/yih-2011-learning-discriminative-projections-text-similarity/)
- [next → A distributional similarity approach to the detection of semantic change in the Google Books Ngram corpus.](https://md.research.elamur.ai/free/en/papers/gulordava-2011-distributional-similarity-approach-detection-semantic/)
- [↑ back to papers of 2011](https://md.research.elamur.ai/free/en/papers/2011/)

## References

14 references, 2 of them with a card of ours. ↗ leads to the work, «find» searches for one we have no id for yet.

1. Peter F. Brown, Vincent J. Della Pietra, Peter V. deSouza, Jenifer C. Lai, and Robert L. Mercer. 1992. Class-based n-gram models of natural language. Computational Linguistics, 18(4):467–479. [find](https://scholar.google.com/scholar?q=Peter+F%2E+Brown%2C+Vincent+J%2E+Della+Pietra%2C+Peter+V%2E+deSouza%2C+Jenifer+C%2E+Lai%2C+and+Robert+L%2E+Mercer%2E+1992%2E+Class%2Dbased+n%2Dgram+models+of+natural+language%2E+Computational+Linguistics%2C+18%284%29%3A467%E2%80%93479%2E)
2. Yee Seng Chan and Dan Roth. 2010. Exploiting background knowledge for relation extraction. In Proceedings of the International Conference on Computational Linguistics (COLING), pages 152–160. [find](https://scholar.google.com/scholar?q=Yee+Seng+Chan+and+Dan+Roth%2E+2010%2E+Exploiting+background+knowledge+for+relation+extraction%2E+In+Proceedings+of+the+International+Conference+on+Computational+Linguistics+%28COLING%29%2C+pages+152%E2%80%93160%2E)
3. Rong-En Fan, Kai-Wei Chang, Cho-Jui Hsieh, Xiang-Rui Wang, and Chih-Jen Lin. 2008. Liblinear: A library for large linear classification. Journal of Machine Learning Research, 9:1871–1874. [find](https://scholar.google.com/scholar?q=Rong%2DEn+Fan%2C+Kai%2DWei+Chang%2C+Cho%2DJui+Hsieh%2C+Xiang%2DRui+Wang%2C+and+Chih%2DJen+Lin%2E+2008%2E+Liblinear%3A+A+library+for+large+linear+classification%2E+Journal+of+Machine+Learning+Research%2C+9%3A1871%E2%80%931874%2E)
4. Katrin Fundel, Robert Küffner, and Ralf Zimmer. 2007. Relex – Relation extraction using dependency parse trees. Bioinformatics, 23(3):365–371. [find](https://scholar.google.com/scholar?q=Katrin+Fundel%2C+Robert+Ku%CC%88ffner%2C+and+Ralf+Zimmer%2E+2007%2E+Relex+%E2%80%93+Relation+extraction+using+dependency+parse+trees%2E+Bioinformatics%2C+23%283%29%3A365%E2%80%93371%2E)
5. Mark A. Greenwood and Mark Stevenson. 2006. Improving semi-supervised acquisition of relation extraction patterns. In Proceedings of the COLING-ACL Workshop on Information Extraction Beyond The Document, pages 29–35. [find](https://scholar.google.com/scholar?q=Mark+A%2E+Greenwood+and+Mark+Stevenson%2E+2006%2E+Improving+semi%2Dsupervised+acquisition+of+relation+extraction+patterns%2E+In+Proceedings+of+the+COLING%2DACL+Workshop+on+Information+Extraction+Beyond+The+Docume)
6. Jing Jiang and ChengXiang Zhai. 2007. A systematic exploration of the feature space for relation extraction. In Proceedings of Human Language Technologies - North American Chapter of the Association for Computational Linguistics (HLT-NAACL), pages 113–120. [find](https://scholar.google.com/scholar?q=Jing+Jiang+and+ChengXiang+Zhai%2E+2007%2E+A+systematic+exploration+of+the+feature+space+for+relation+extraction%2E+In+Proceedings+of+Human+Language+Technologies+%2D+North+American+Chapter+of+the+Association+f)
7. Jing Jiang. 2009. Multi-task transfer learning for weakly-supervised relation extraction. In Proceedings of the Annual Meeting of the Association for Computational Linguistics and International Joint Conference on Natural Language Processing (ACL-IJCNLP), pages 1012–1020. [find](https://scholar.google.com/scholar?q=Jing+Jiang%2E+2009%2E+Multi%2Dtask+transfer+learning+for+weakly%2Dsupervised+relation+extraction%2E+In+Proceedings+of+the+Annual+Meeting+of+the+Association+for+Computational+Linguistics+and+International+Joint+)
8. Nanda Kambhatla. 2004. Combining lexical, syntactic, and semantic features with maximum entropy models for information extraction. In Proceedings of the Annual Meeting of the Association for Computational Linguistics (ACL), pages 178–181. [find](https://scholar.google.com/scholar?q=Nanda+Kambhatla%2E+2004%2E+Combining+lexical%2C+syntactic%2C+and+semantic+features+with+maximum+entropy+models+for+information+extraction%2E+In+Proceedings+of+the+Annual+Meeting+of+the+Association+for+Computati)
9. Dan Klein and Christoper D. Manning. 2003. Fast exact inference with a factored model for natural language parsing. In The Conference on Advances in Neural Information Processing Systems (NIPS), pages 3–10. [find](https://scholar.google.com/scholar?q=Dan+Klein+and+Christoper+D%2E+Manning%2E+2003%2E+Fast+exact+inference+with+a+factored+model+for+natural+language+parsing%2E+In+The+Conference+on+Advances+in+Neural+Information+Processing+Systems+%28NIPS%29%2C+pages)
10. Longhua Qian, Guodong Zhou, Qiaomin Zhu, and Peide Qian. 2008. Relation extraction using convolution tree kernel expanded with entity features. In Pacific Asia Conference on Language, Information and Computation, pages 415–421. [find](https://scholar.google.com/scholar?q=Longhua+Qian%2C+Guodong+Zhou%2C+Qiaomin+Zhu%2C+and+Peide+Qian%2E+2008%2E+Relation+extraction+using+convolution+tree+kernel+expanded+with+entity+features%2E+In+Pacific+Asia+Conference+on+Language%2C+Information+and+)
11. Lev Ratinov and Dan Roth. 2009. [Design challenges and misconceptions in named entity recognition](https://md.research.elamur.ai/free/en/papers/ratinov-2009-design-challenges-misconceptions-named-entity/). In Proceedings of the Annual Conference on Computational Natural Language Learning (CoNLL), pages 147–155.
12. Dan Roth and Wen Tau Yih. 2007. Global inference for entity and relation identification via a linear programming formulation. In Lise Getoor and Ben Taskar, editors, Introduction to Statistical Relational Learning. MIT Press. [find](https://scholar.google.com/scholar?q=Dan+Roth+and+Wen+Tau+Yih%2E+2007%2E+Global+inference+for+entity+and+relation+identification+via+a+linear+programming+formulation%2E+In+Lise+Getoor+and+Ben+Taskar%2C+editors%2C+Introduction+to+Statistical+Relati)
13. Guodong Zhou, Jian Su, Jie Zhang, and Min Zhang. 2005. [Exploring various knowledge in relation extraction](https://md.research.elamur.ai/free/en/papers/zhou-2005-exploring-various-knowledge-relation-extraction/). In Proceedings of the Annual Meeting of the Association for Computational Linguistics (ACL), pages 427–434.
14. Jun Zhu, Zaiqing Nie, Xiaojiang Liu, Bo Zhang, and Ji-Rong Wen. 2009. Statsnowball: a statistical approach to extracting entity relationships. In The International World Wide Web Conference, pages 101–110. [find](https://scholar.google.com/scholar?q=Jun+Zhu%2C+Zaiqing+Nie%2C+Xiaojiang+Liu%2C+Bo+Zhang%2C+and+Ji%2DRong+Wen%2E+2009%2E+Statsnowball%3A+a+statistical+approach+to+extracting+entity+relationships%2E+In+The+International+World+Wide+Web+Conference%2C+pages+101)

*Updated 24 October 2026*