---
title: "Every Child Should Have Parents — paper by Aly et al. (2019)"
description: "Aly et al. propose a simple, parameter-free post-processing pipeline that repairs such taxonomies… 33 references, cited by 16 papers here."
authors: ["Rami Aly", "Shantanu Acharya", "Alexander Ossa", "Arne Köhn", "Chris Biemann", "Alexander Panchenko"]
year: 2019
venue: "Proceedings of the 57th Annual Meeting of the Association for Computational Linguistics"
source: "https://aclanthology.org/P19-1474/"
doi: "10.18653/v1/p19-1474"
arxiv: "1906.02002"
licence: "CC BY 4.0 (ACL Anthology)"
licence_url: "https://creativecommons.org/licenses/by/4.0/"
summary_licence: "Elamur Cosmic License"
canonical: "https://research.elamur.ai/free/en/papers/aly-2019-every-child-should-have-parents/"
---
[Elamur Research](https://md.research.elamur.ai/free/en/) · [this page in HTML](https://research.elamur.ai/free/en/papers/aly-2019-every-child-should-have-parents/)

[home](https://md.research.elamur.ai/free/en/) / [papers](https://md.research.elamur.ai/free/en/papers/) / [2019](https://md.research.elamur.ai/free/en/papers/2019/page/6/) / Every Child Should Have Parents: A Taxonomy Refinement Algorithm Based on Hyperbolic Term Embeddings

# Every Child Should Have Parents: A Taxonomy Refinement Algorithm Based on Hyperbolic Term Embeddings

**Authors:** [Rami Aly](https://md.research.elamur.ai/free/en/authors/rami-aly/), [Shantanu Acharya](https://md.research.elamur.ai/free/en/authors/shantanu-acharya/), Alexander Ossa, [Arne Köhn](https://md.research.elamur.ai/free/en/authors/arne-kohn/), [Chris Biemann](https://md.research.elamur.ai/free/en/authors/chris-biemann/) and [Alexander Panchenko](https://md.research.elamur.ai/free/en/authors/alexander-panchenko/) · **Year:** 2019 · **Venue:** [Proceedings of the 57th Annual Meeting of the Association for Computational Linguistics](https://md.research.elamur.ai/free/en/venues/acl/) ([ACL 2019](https://md.research.elamur.ai/free/en/venues/acl/2019/)) · **Publisher:** [Association for Computational Linguistics](https://md.research.elamur.ai/free/en/publishers/association-for-computational-linguistics/) · **Licence:** [CC BY 4.0 (ACL Anthology)](https://creativecommons.org/licenses/by/4.0/)

**Links:** [Publisher's page](https://aclanthology.org/P19-1474/) · [PDF](https://aclanthology.org/P19-1474.pdf) · [DOI 10.18653/v1/p19-1474](https://doi.org/10.18653/v1/p19-1474) · [arXiv 1906.02002](https://arxiv.org/abs/1906.02002) · [OpenAlex](https://openalex.org/W2948074670)

**Citations:** 46 citations (OpenAlex, 2026-10-14) · [16](https://md.research.elamur.ai/free/en/papers/aly-2019-every-child-should-have-parents/cited-by/) in our library, of the 149,439 papers whose bibliographies we read

## Summary
Taxonomies built automatically from text often contain two kinds of errors: orphans, which are terms left with no connections, and outliers, which are terms attached to the wrong parent. Aly et al. propose a simple, parameter-free post-processing pipeline that repairs such taxonomies using hyperbolic Poincaré embeddings. They first train domain-specific Poincaré embeddings on noisy IS-A pairs. These pairs are mined with lexical-syntactic patterns (PattaMaika, PatternSim, WebISA) from large general corpora and from crawled domain-specific corpora. The embeddings are then used to rank parent–child links. Links ranked worse than the mean are dropped, and orphans are attached to their nearest plausible parent. Leftover compound terms are linked to a parent contained within them as a substring, and cycles are removed with the Tarjan algorithm. On SemEval-2016 Task 13 (TExEval), the method improves the outputs of three participating systems, with the largest gains for TAXI. It also beats a word2vec CBOW variant and an embedding trained on WordNet. The work shows that hierarchy-aware embeddings can be learned directly from crawled text, with no curated lexical database required.

## Key points
- The method was tested on outputs of TAXI, USAAR and JUNLP across the environment, science and food domains. Poincaré-based refinement gave statistically significant gains under the McNemar test.
- English F1 for TAXI improved in all three domains: environment rose from 26.9 to 30.9, science from 36.7 to 41.4, and food from 27.9 to 34.1.
- The embeddings were trained on noisy relations from 59.2 GB of general text (Wikipedia, Gigaword, ukWac, LCC) plus BootCat/focused-crawl domain corpora. Pairs seen fewer than 3 times (domain corpora) or 5 times (general corpora) were discarded.
- The WordNet-based Poincaré model (P-WN) was more accurate but covered fewer terms, especially multiword terms such as "second language acquisition". As a result, the domain-specific model attached more orphans.
- Results for French, Italian and Dutch were mixed. Food taxonomies improved in every language, while environment declined for Italian and Dutch because too few relations were extracted. Code is available at github.com/uhh-lt/Taxonomy_Refinement_Embeddings.

*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 Cambridge](https://md.research.elamur.ai/free/en/universities/university-of-cambridge/) — University, Cambridge, United Kingdom · 287 works · 125,174 citations
- [Saarland University](https://md.research.elamur.ai/free/en/universities/saarland-university/) — University, Saarbrücken, Germany · 153 works · 18,502 citations
- [Universität Hamburg](https://md.research.elamur.ai/free/en/universities/universitat-hamburg/) — University, Hamburg, Germany · 84 works · 5,692 citations
- [Skolkovo Institute of Science and Technology](https://md.research.elamur.ai/free/en/universities/skolkovo-institute-of-science-and-technology/) — University, Skolkovo, Russia · 6 works · 204 citations

📚 Shelves (4): [🖥️ Computer Science](https://md.research.elamur.ai/free/en/shelves/computer-science/) · 19,503 works · [🤖 Artificial Intelligence](https://md.research.elamur.ai/free/en/shelves/computer-science/artificial-intelligence/) · 18,805 works · [🎭 Arts and Humanities](https://md.research.elamur.ai/free/en/shelves/arts-and-humanities/) · 17,959 works · [🗣️ Language and Linguistics](https://md.research.elamur.ai/free/en/shelves/arts-and-humanities/language-and-linguistics/) · 17,789 works

## Cite

Cite as: Rami Aly, Shantanu Acharya, Alexander Ossa, Arne Köhn, Chris Biemann and Alexander Panchenko. 2019. Every Child Should Have Parents: A Taxonomy Refinement Algorithm Based on Hyperbolic Term Embeddings. Proceedings of the 57th Annual Meeting of the Association for Computational Linguistics.

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

```bibtex
@inproceedings{aly2019every,
  title = {{Every Child Should Have Parents: A Taxonomy Refinement Algorithm Based on Hyperbolic Term Embeddings}},
  author = {Rami Aly, Shantanu Acharya, Alexander Ossa, Arne Köhn, Chris Biemann and Alexander Panchenko},
  booktitle = {Proceedings of the 57th Annual Meeting of the Association for Computational Linguistics},
  year = {2019},
  doi = {10.18653/v1/p19-1474},
  eprint = {1906.02002},
  archivePrefix = {arXiv},
  url = {https://doi.org/10.18653/v1/p19-1474}
}
```

- [← previous A Logic-Driven Framework for Consistency of Neural Models](https://md.research.elamur.ai/free/en/papers/li-2019-logic-driven-framework-consistency-neural/)
- [next → ArSentD-LEV: A Multi-Topic Corpus for Target-based Sentiment Analysis in Arabic Levantine Tweets](https://md.research.elamur.ai/free/en/papers/baly-2019-arsentd-lev-multi-topic-corpus/)
- [↑ back to papers of 2019](https://md.research.elamur.ai/free/en/papers/2019/page/6/)

## References

33 references, 12 of them with a card of ours. ↗ leads to the work, «find» searches for one we have no id for yet.

1. Marco Baroni and Silvia Bernardini. 2004. [Bootcat: Bootstrapping corpora and terms from the web](https://md.research.elamur.ai/free/en/papers/baroni-2004-bootcat-bootstrapping-corpora-terms-web/). In Proceedings of the Fourth International Conference on Language Resources and Evaluation (LREC’04), pages 1313–1316, Lisbon, Portugal.
2. Chris Biemann. 2005. Ontology learning from text: A survey of methods. LDV Forum, 20(2):75–93. [find](https://scholar.google.com/scholar?q=Chris+Biemann%2E+2005%2E+Ontology+learning+from+text%3A+A+survey+of+methods%2E+LDV+Forum%2C+20%282%29%3A75%E2%80%9393%2E)
3. Georgeta Bordea, Paul Buitelaar, Stefano Faralli, and Roberto Navigli. 2015. Semeval-2015 task 17: Taxonomy Extraction Evaluation (TExEval). In Proceedings of the 9th International Workshop on Semantic Evaluation (SemEval 2015), pages 902–910, Denver, CO, USA. [find](https://scholar.google.com/scholar?q=Georgeta+Bordea%2C+Paul+Buitelaar%2C+Stefano+Faralli%2C+and+Roberto+Navigli%2E+2015%2E+Semeval%2D2015+task+17%3A+Taxonomy+Extraction+Evaluation+%28TExEval%29%2E+In+Proceedings+of+the+9th+International+Workshop+on+Semanti)
4. Georgeta Bordea, Els Lefever, and Paul Buitelaar. 2016. SemEval-2016 Task 13: Taxonomy Extraction Evaluation (TExEval-2). In Proceedings of the 10th International Workshop on Semantic Evaluation (SemEval-2016), pages 1081–1091, San Diego, CA, USA. [find](https://scholar.google.com/scholar?q=Georgeta+Bordea%2C+Els+Lefever%2C+and+Paul+Buitelaar%2E+2016%2E+SemEval%2D2016+Task+13%3A+Taxonomy+Extraction+Evaluation+%28TExEval%2D2%29%2E+In+Proceedings+of+the+10th+International+Workshop+on+Semantic+Evaluation+%28SemE)
5. Adriano Ferraresi, Eros Zanchetta, Marco Baroni, and Silvia Bernardini. 2008. Introducing and evaluating ukWaC, a very large web-derived corpus of English. In Proceedings of the 4th Web as Corpus Workshop. Can we beat Google?, pages 47–54, Marrakech, Morocco. [find](https://scholar.google.com/scholar?q=Adriano+Ferraresi%2C+Eros+Zanchetta%2C+Marco+Baroni%2C+and+Silvia+Bernardini%2E+2008%2E+Introducing+and+evaluating+ukWaC%2C+a+very+large+web%2Dderived+corpus+of+English%2E+In+Proceedings+of+the+4th+Web+as+Corpus+Work)
6. Ruiji Fu, Jiang Guo, Bing Qin, Wanxiang Che, Haifeng Wang, and Ting Liu. 2014. [Learning semantic hierarchies via word embeddings](https://md.research.elamur.ai/free/en/papers/fu-2014-learning-semantic-hierarchies-word-embeddings/). In Proceedings of the 52nd Annual Meeting of the Association for Computational Linguistics, volume 1, pages 1199–1209, Baltimore, MD, USA.
7. Conference on Computer Vision and Pattern Recognition (CVPR). [find](https://scholar.google.com/scholar?q=Conference+on+Computer+Vision+and+Pattern+Recognition+%28CVPR%29%2E)
8. Peter Kluegl, Martin Toepfer, Philip-Daniel Beck, Georg Fette, and Frank Puppe. 2016. UIMA Ruta: Rapid development of rule-based information extraction applications. Natural Language Engineering, 22(1):1–40. [find](https://scholar.google.com/scholar?q=Peter+Kluegl%2C+Martin+Toepfer%2C+Philip%2DDaniel+Beck%2C+Georg+Fette%2C+and+Frank+Puppe%2E+2016%2E+UIMA+Ruta%3A+Rapid+development+of+rule%2Dbased+information+extraction+applications%2E+Natural+Language+Engineering%2C+22%281)
9. Matt Le, Stephen Roller, Laetitia Papaxanthos, Douwe Kiela, and Maximilian Nickel. 2019. Inferring concept hierarchies from text corpora via hyperbolic embeddings. arXiv preprint arXiv:1902.00913. [find](https://scholar.google.com/scholar?q=Matt+Le%2C+Stephen+Roller%2C+Laetitia+Papaxanthos%2C+Douwe+Kiela%2C+and+Maximilian+Nickel%2E+2019%2E+Inferring+concept+hierarchies+from+text+corpora+via+hyperbolic+embeddings%2E+arXiv+preprint+arXiv%3A1902%2E00913%2E)
10. Omer Levy, Steffen Remus, Chris Biemann, and Ido Dagan. 2015. [Do supervised distributional methods really learn lexical inference relations?](https://md.research.elamur.ai/free/en/papers/levy-2015-do-supervised-distributional-methods-really/) In Proceedings of the 2015 Conference of the North American Chapter of the Association for Computational Linguistics: Human Language Technologies, pages 970–97…
11. Promita Maitra and Dipankar Das. 2016. JUNLP at SemEval-2016 task 13: A language independent approach for hypernym identification. In Proceedings of the 10th International Workshop on Semantic Evaluation (SemEval-2016), pages 1310–1314, San Diego, CA, USA. [find](https://scholar.google.com/scholar?q=Promita+Maitra+and+Dipankar+Das%2E+2016%2E+JUNLP+at+SemEval%2D2016+task+13%3A+A+language+independent+approach+for+hypernym+identification%2E+In+Proceedings+of+the+10th+International+Workshop+on+Semantic+Evaluat)
12. Quinn McNemar. 1947. Note on the sampling error of the difference between correlated proportions or percentages. Psychometrika, 12(2):153–157. [find](https://scholar.google.com/scholar?q=Quinn+McNemar%2E+1947%2E+Note+on+the+sampling+error+of+the+difference+between+correlated+proportions+or+percentages%2E+Psychometrika%2C+12%282%29%3A153%E2%80%93157%2E)
13. Tomas Mikolov, Kai Chen, G.S Corrado, and Jeffrey Dean. 2013a. [Efficient estimation of word representations in vector space](https://md.research.elamur.ai/free/en/papers/mikolov-2013-efficient-estimation-word-representations-vector/). In ICLR Workshop, Scottsdale, AZ, USA.
14. Dirk Goldhahn, Thomas Eckart, and Uwe Quasthoff. 2012. [Building large monolingual dictionaries at the Leipzig Corpora Collection: From 100 to 200 languages](https://md.research.elamur.ai/free/en/papers/goldhahn-2012-building-large-monolingual-dictionaries-leipzig/). In Proceedings of the Eight International Conference on Language Resources and Evaluation, pages 759–765, Istanbul, Turkey.
15. Tomas Mikolov, Ilya Sutskever, Kai Chen, Greg S. Corrado, and Jeff Dean. 2013b. [Distributed representations of words and phrases and their compositionality](https://md.research.elamur.ai/free/en/papers/mikolov-2013-distributed-representations-words-phrases-their/). In Advances in neural information processing systems, pages 3111–3119, Stateline, NV, USA.
16. Gregory Grefenstette. 2015. INRIASAC: Simple Hypernym Extraction Methods. In Proceedings of the 9th International Workshop on Semantic Evaluation (SemEval 2015), pages 911–914, Denver, CO, USA. [find](https://scholar.google.com/scholar?q=Gregory+Grefenstette%2E+2015%2E+INRIASAC%3A+Simple+Hypernym+Extraction+Methods%2E+In+Proceedings+of+the+9th+International+Workshop+on+Semantic+Evaluation+%28SemEval+2015%29%2C+pages+911%E2%80%93914%2C+Denver%2C+CO%2C+USA%2E)
17. Maximillian Nickel and Douwe Kiela. 2017. [Poincaré embeddings for learning hierarchical representations](https://md.research.elamur.ai/free/en/papers/nickel-2017-poincare-embeddings-learning-hierarchical-representations/). In Advances in Neural Information Processing Systems 30, pages 6338–6347, Long Beach, CA, USA.
18. Marti A. Hearst. 1992. [Automatic acquisition of hyponyms from large text corpora](https://md.research.elamur.ai/free/en/papers/hearst-1992-automatic-acquisition-hyponyms-large-text/). In Proceedings of the 14th Conference on Computational Linguistics, pages 539–545, Nantes, France.
19. Kris Heylen, Yves Peirsman, Dirk Geeraerts, and Dirk Speelman. 2008. Modelling word similarity: an evaluation of automatic synonymy extraction algorithms. In Proceedings of the sixth international language resources and evaluation, pages 3243– 3249, Marrakech, Morocco. [find](https://scholar.google.com/scholar?q=Kris+Heylen%2C+Yves+Peirsman%2C+Dirk+Geeraerts%2C+and+Dirk+Speelman%2E+2008%2E+Modelling+word+similarity%3A+an+evaluation+of+automatic+synonymy+extraction+algorithms%2E+In+Proceedings+of+the+sixth+international+lan)
20. Valentin Khrulkov, Leyla Mirvakhabova, Evgeniya Ustinova, Ivan Oseledets, and Victor Lempitsky. 2019. Hyperbolic image embeddings. In The IEEE Alexander Panchenko, Stefano Faralli, Eugen Ruppert, Steffen Remus, Hubert Naets, Cédrick Fairon, Simone P. Ponzetto, and Chris Biemann. 2016. TAXI at SemEv… [find](https://scholar.google.com/scholar?q=Valentin+Khrulkov%2C+Leyla+Mirvakhabova%2C+Evgeniya+Ustinova%2C+Ivan+Oseledets%2C+and+Victor+Lempitsky%2E+2019%2E+Hyperbolic+image+embeddings%2E+In+The+IEEE+Alexander+Panchenko%2C+Stefano+Faralli%2C+Eugen+Ruppert%2C+Stef)
21. Alexander Panchenko, Olga Morozova, and Hubert Naets. 2012. A semantic similarity measure based on lexico-syntactic patterns. In Proceedings of KONVENS 2012, pages 174–178, Vienna, Austria. [find](https://scholar.google.com/scholar?q=Alexander+Panchenko%2C+Olga+Morozova%2C+and+Hubert+Naets%2E+2012%2E+A+semantic+similarity+measure+based+on+lexico%2Dsyntactic+patterns%2E+In+Proceedings+of+KONVENS+2012%2C+pages+174%E2%80%93178%2C+Vienna%2C+Austria%2E)
22. Robert Parker, David Graff, Junbo Kong, Ke Chen, and Kazuaki Maeda. 2009. English gigaword forth edition. In Linguistic Data Consortium, Philadelphia, PA, USA. [find](https://scholar.google.com/scholar?q=Robert+Parker%2C+David+Graff%2C+Junbo+Kong%2C+Ke+Chen%2C+and+Kazuaki+Maeda%2E+2009%2E+English+gigaword+forth+edition%2E+In+Linguistic+Data+Consortium%2C+Philadelphia%2C+PA%2C+USA%2E)
23. Joel Pocostales. 2016. NUIG-UNLP at SemEval-2016 Task 13: A Simple Word Embedding-based Approach for Taxonomy Extraction. In Proceedings of the 10th International Workshop on Semantic Evaluation (SemEval-2016), pages 1298–1302, San Diego, CA, USA. [find](https://scholar.google.com/scholar?q=Joel+Pocostales%2E+2016%2E+NUIG%2DUNLP+at+SemEval%2D2016+Task+13%3A+A+Simple+Word+Embedding%2Dbased+Approach+for+Taxonomy+Extraction%2E+In+Proceedings+of+the+10th+International+Workshop+on+Semantic+Evaluation+%28SemE)
24. Chao Zhang, Fangbo Tao, Xiusi Chen, Jiaming Shen, Meng Jiang, Brian Sadler, Michelle Vanni, and Jiawei Han. 2018. TaxoGen: Unsupervised topic taxonomy construction by adaptive term embedding and clustering. In Proceedings of the 24th ACM SIGKDD International Conference on Knowledge Discovery & Data… [find](https://scholar.google.com/scholar?q=Chao+Zhang%2C+Fangbo+Tao%2C+Xiusi+Chen%2C+Jiaming+Shen%2C+Meng+Jiang%2C+Brian+Sadler%2C+Michelle+Vanni%2C+and+Jiawei+Han%2E+2018%2E+TaxoGen%3A+Unsupervised+topic+taxonomy+construction+by+adaptive+term+embedding+and+clust)
25. Steffen Remus and Chris Biemann. 2016. Domain-specific corpus expansion with focused webcrawling. In Proceedings of the 10th International Conference on Language Resources and Evaluation, pages 3607–3611, Portorož, Slovenia. [find](https://scholar.google.com/scholar?q=Steffen+Remus+and+Chris+Biemann%2E+2016%2E+Domain%2Dspecific+corpus+expansion+with+focused+webcrawling%2E+In+Proceedings+of+the+10th+International+Conference+on+Language+Resources+and+Evaluation%2C+pages+3607%E2%80%933)
26. Julian Seitner, Christian Bizer, Kai Eckert, Stefano Faralli, Robert Meusel und Heiko Paulheim, and Simone P. Ponzetto. 2016. [A large database of hypernymy relations extracted from the Web](https://md.research.elamur.ai/free/en/papers/seitner-2016-large-database-hypernymy-relations-extracted/). In Proceedings of the 10th International Conference on Language Resources and Evaluation, pages 360–367, Porto…
27. Vered Shwartz, Yoav Goldberg, and Ido Dagan. 2016. [Improving hypernymy detection with an integrated path-based and distributional method](https://md.research.elamur.ai/free/en/papers/shwartz-2016-improving-hypernymy-detection-integrated-path/). In Proceedings of the 54th Annual Meeting of the Association for Computational Linguistics, pages 2389–2398, Berlin, Germany.
28. John Stillwell. 1996. Sources of hyperbolic geometry. History of Mathematics, Volume 10. American Mathematical Society. [find](https://scholar.google.com/scholar?q=John+Stillwell%2E+1996%2E+Sources+of+hyperbolic+geometry%2E+History+of+Mathematics%2C+Volume+10%2E+American+Mathematical+Society%2E)
29. Liling Tan, Francis Bond, and Josef van Genabith. 2016. USAAR at SemEval-2016 task 13: Hyponym endocentricity. In Proceedings of the 10th International Workshop on Semantic Evaluation (SemEval-2016), pages 1303–1309, San Diego, CA, USA. [find](https://scholar.google.com/scholar?q=Liling+Tan%2C+Francis+Bond%2C+and+Josef+van+Genabith%2E+2016%2E+USAAR+at+SemEval%2D2016+task+13%3A+Hyponym+endocentricity%2E+In+Proceedings+of+the+10th+International+Workshop+on+Semantic+Evaluation+%28SemEval%2D2016%29%2C+)
30. Liling Tan, Rohit Gupta, and Josef van Genabith. 2015. USAAR-WLV: Hypernym generation with Deep Neural Nets. In Proceedings of the 9th International Workshop on Semantic Evaluation (SemEval 2015), pages 932–937, Denver, CO, USA. [find](https://scholar.google.com/scholar?q=Liling+Tan%2C+Rohit+Gupta%2C+and+Josef+van+Genabith%2E+2015%2E+USAAR%2DWLV%3A+Hypernym+generation+with+Deep+Neural+Nets%2E+In+Proceedings+of+the+9th+International+Workshop+on+Semantic+Evaluation+%28SemEval+2015%29%2C+pag)
31. Robert Tarjan. 1972. Depth first search and linear graph algorithms. SIAM Journal on Computing, 1(2):146 –160. [find](https://scholar.google.com/scholar?q=Robert+Tarjan%2E+1972%2E+Depth+first+search+and+linear+graph+algorithms%2E+SIAM+Journal+on+Computing%2C+1%282%29%3A146+%E2%80%93160%2E)
32. Chengyu Wang, Xiaofeng He, and Aoying Zhou. 2017. [A short survey on taxonomy learning from text corpora: Issues, resources and recent advances](https://md.research.elamur.ai/free/en/papers/wang-2017-short-survey-taxonomy-learning-text/). In Proceedings of the 2017 Conference on Empirical Methods in Natural Language Processing, pages 1190– 1203, Copenhagen, Denmark.
33. Julie Weeds, Daoud Clarke, Jeremy Reffin, David Weir, and Bill Keller. 2014. [Learning to distinguish hypernyms and co-hyponyms](https://md.research.elamur.ai/free/en/papers/weeds-2014-learning-distinguish-hypernyms-co-hyponyms/). In Proceedings of COLING 2014, the 25th International Conference on Computational Linguistics, pages 2249– 2259, Dublin, Ireland.

*Updated 3 November 2026*