---
title: "Using Encyclopedic Knowledge for Named entity — Bunescu 2006"
description: "The paper addresses named entity detection and disambiguation in open-domain text."
authors: ["Razvan Bunescu", "Marius Paşca"]
year: 2006
venue: "11th Conference of the European Chapter of the Association for Computational Linguistics"
source: "https://aclanthology.org/E06-1002/"
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/bunescu-2006-encyclopedic-knowledge-named-entity-disambiguation/"
---
[Elamur Research](https://md.research.elamur.ai/free/en/) · [this page in HTML](https://research.elamur.ai/free/en/papers/bunescu-2006-encyclopedic-knowledge-named-entity-disambiguation/)

[home](https://md.research.elamur.ai/free/en/) / [papers](https://md.research.elamur.ai/free/en/papers/) / [2006](https://md.research.elamur.ai/free/en/papers/2006/) / Using Encyclopedic Knowledge for Named entity Disambiguation

# Using Encyclopedic Knowledge for Named entity Disambiguation

**Authors:** [Razvan Bunescu](https://md.research.elamur.ai/free/en/authors/razvan-bunescu/), [Marius Paşca](https://md.research.elamur.ai/free/en/authors/marius-pasca/) · **Year:** 2006 · **Venue:** 11th Conference of the European Chapter of the 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/E06-1002/) · [PDF](https://aclanthology.org/E06-1002.pdf) · [OpenAlex](https://openalex.org/W1548663377)

**Citations:** 793 citations (openalex-snapshot · 2026-10-14) · [185](#cited-by) in our library, of the 140884 papers whose bibliographies we read

## Summary
The paper addresses named entity detection and disambiguation in open-domain text. Given a proper name in context, the system decides whether it refers to an entity known to a dictionary and, if so, which of the candidates sharing that name is meant. The motivating use case is web search, where queries such as "John Williams" may point to a composer, a wrestler or a soldier. The authors build the name dictionary from the English Wikipedia (May 2005) using article titles, redirect pages and disambiguation pages. Wikipedia hyperlinks serve as automatically labeled training examples. A ranking SVM with a "taxonomy kernel" learns associations between words around the name and the Wikipedia categories of candidate entities. Its baseline is unsupervised cosine similarity between the context and the article text. Across four scenarios built on the People by occupation category tree, the taxonomy kernel clearly beats cosine similarity. The gain is smallest when detection of out-of-Wikipedia entities is added. The work matters because it shows a collaboratively built encyclopedia can act as a large, structured resource for entity linking.

## Key points
- Resource: the English Wikipedia from May 2005, with 577,860 articles and 59,759 categories. Capitalization heuristics on titles identify close to half a million named entities.
- Dictionary entries combine each entity's title, its redirect names and the disambiguation pages that list it. Each name maps to the set of entities it can denote.
- The query context is a 55-word window centered on the name. This size was taken from earlier cross-document coreference work.
- Taxonomy kernel (TK) vs. cosine (Cos) accuracy:
  - S1, 110 top-level occupation categories: 77.2% vs. 61.5%
  - S2, 540 categories: 68.4% vs. 55.8%
  - S3, 2,847 categories: 68.0% vs. 55.4%
  - S4, detection plus disambiguation with 10% out-of-Wikipedia answers: 84.8% vs. 82.3%
- Training and test sets share no query names. Each scenario's training was capped at about three days on a single Pentium 4.

## Interesting ideas
- **Hyperlinks as free supervision:** every link in an encyclopedia pairs a surface string with its true referent, which yields labeled disambiguation data at no annotation cost. The same trick could work for any wiki-style resource, including smaller-language editions of Wikipedia.
- **Aliases from redirects and disambiguation pages:** redirects capture full names, acronyms, misspellings and translations, while disambiguation pages list every referent of an ambiguous name. Together they give a ready-made many-to-many name–entity lexicon.
- **Context words matched to categories, not to article text:** learning which context words go with which taxonomy categories lets the model generalize to entities whose own article shares few words with the query. This fits any domain with a category hierarchy.
- **Simulating unknown entities:** in S4, some candidates are relabeled as "out-of-Wikipedia" so that the model must also decide when no known entity fits. This is a cheap way to test the NIL/detection case without extra annotation.
- **Splitting data by name:** keeping all queries for a given name on one side of the train/test split stops the model from memorizing rare, answer-specific words such as album or book titles. It gives a fairer comparison with unsupervised baselines and is good practice for any entity-centric evaluation.

*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: [Natural language processing](https://md.research.elamur.ai/free/en/shelves/nlp/)

Topics: [Named entity recognition](https://md.research.elamur.ai/free/en/topics/named-entity-recognition/)

## Cited by

Cited by

185 papers of our library cite it (of the 140884 whose bibliographies we read); 2 of them have a page here, 183 not yet.

- [Entity Linking in 100 Languages](https://md.research.elamur.ai/free/en/papers/botha-2020-entity-linking-100-languages/)
- [Joint Learning of the Embedding of Words and Entities for Named Entity Disambiguation](https://md.research.elamur.ai/free/en/papers/yamada-2016-joint-learning-embedding-words-entities/)

## Cite

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

```bibtex
@inproceedings{bunescu2006using,
  title = {{Using Encyclopedic Knowledge for Named entity Disambiguation}},
  author = {Bunescu, Razvan and Paşca, Marius},
  booktitle = {11th Conference of the European Chapter of the Association for Computational Linguistics},
  year = {2006},
  url = {https://aclanthology.org/E06-1002/}
}
```

- [← previous Unsupervised Multilingual Sentence Boundary Detection](https://md.research.elamur.ai/free/en/papers/cl-2006-kiss-unsupervised-multilingual-sentence-boundary-detection/)
- [next → Using Lexical Dependency and Ontological Knowledge to Improve a Detailed Syntactic and Semantic Tagger of English](https://md.research.elamur.ai/free/en/papers/finch-2006-lexical-dependency-ontological-knowledge-improve/)
- [↑ back to papers of 2006](https://md.research.elamur.ai/free/en/papers/2006/)

## References

10 references, 1 of them with a card of ours. ↗ leads to the work, «find» searches for one we have no id for yet.

1. Ricardo Baeza-Yates and Berthier Ribeiro-Neto. 1999. Modern Information Retrieval. ACM Press, New York. [find](https://scholar.google.com/scholar?q=Ricardo+Baeza%2DYates+and+Berthier+Ribeiro%2DNeto%2E+1999%2E+Modern+Information+Retrieval%2E+ACM+Press%2C+New+York%2E)
2. Massimiliano Ciaramita, Thomas Hofmann, and Mark Johnson. 2003. Hierarchical semantic classification: Word sense disambiguation with world knowledge. In The 18th International Joint Conference on Artificial Intelligence, Acapulco, Mexico. [find](https://scholar.google.com/scholar?q=Massimiliano+Ciaramita%2C+Thomas+Hofmann%2C+and+Mark+Johnson%2E+2003%2E+Hierarchical+semantic+classification%3A+Word+sense+disambiguation+with+world+knowledge%2E+In+The+18th+International+Joint+Conference+on+Arti)
3. Robert Dale. 2003. Computational linguistics. Special Issue on the Web as a Corpus, 29(3), September. [find](https://scholar.google.com/scholar?q=Robert+Dale%2E+2003%2E+Computational+linguistics%2E+Special+Issue+on+the+Web+as+a+Corpus%2C+29%283%29%2C+September%2E)
4. Gottlob Frege. 1999. On sense and reference. In Maria Baghramian, editor, Modern Philosophy of Language, pages 3–25. Counterpoint Press. [find](https://scholar.google.com/scholar?q=Gottlob+Frege%2E+1999%2E+On+sense+and+reference%2E+In+Maria+Baghramian%2C+editor%2C+Modern+Philosophy+of+Language%2C+pages+3%E2%80%9325%2E+Counterpoint+Press%2E)
5. Chung Heong Gooi and James Allan. 2004. Cross-document coreference on a large scale corpus. In Proceedings of Human Language Technology Conference / North American Association for Computational Linguistics Annual Meeting, Boston, MA. [find](https://scholar.google.com/scholar?q=Chung+Heong+Gooi+and+James+Allan%2E+2004%2E+Cross%2Ddocument+coreference+on+a+large+scale+corpus%2E+In+Proceedings+of+Human+Language+Technology+Conference+%2F+North+American+Association+for+Computational+Lingui) [our card](https://md.research.elamur.ai/free/en/papers/gooi-2004-cross-document-coreference-large-scale/)
6. Thorsten Joachims. 1999. Making large-scale SVM learning practical. In B. Schölkopf, C. J. C. Burges, and A. J. Smola, editors, Advances in Kernel Methods - Support Vector Learning, pages 169–184. MIT Press. [find](https://scholar.google.com/scholar?q=Thorsten+Joachims%2E+1999%2E+Making+large%2Dscale+SVM+learning+practical%2E+In+B%2E+Scho%CC%88lkopf%2C+C%2E+J%2E+C%2E+Burges%2C+and+A%2E+J%2E+Smola%2C+editors%2C+Advances+in+Kernel+Methods+%2D+Support+Vector+Learning%2C+pages+169%E2%80%93184%2E+MI)
7. Thorsten Joachims. 2002. Optimizing search engines using clickthrough data. In Proceedings of the eighth ACM SIGKDD international conference on Knowledge discovery and data mining, pages 133–142. [find](https://scholar.google.com/scholar?q=Thorsten+Joachims%2E+2002%2E+Optimizing+search+engines+using+clickthrough+data%2E+In+Proceedings+of+the+eighth+ACM+SIGKDD+international+conference+on+Knowledge+discovery+and+data+mining%2C+pages+133%E2%80%93142%2E)
8. Andrew McCallum, R. Rosenfeld, Tom Mitchell, and A. Y. Ng. 1998. Improving text classification by shrinkage in a hierarchy of classes. In Proceedings of the Fifteenth International Conference on Machine Learning (ICML-98), pages 359–367, Madison, WI. [find](https://scholar.google.com/scholar?q=Andrew+McCallum%2C+R%2E+Rosenfeld%2C+Tom+Mitchell%2C+and+A%2E+Y%2E+Ng%2E+1998%2E+Improving+text+classification+by+shrinkage+in+a+hierarchy+of+classes%2E+In+Proceedings+of+the+Fifteenth+International+Conference+on+Machi)
9. M. Remy. 2002. Wikipedia: The free encyclopedia. Online Information Review, 26(6):434. www.wikipedia.org. [find](https://scholar.google.com/scholar?q=M%2E+Remy%2E+2002%2E+Wikipedia%3A+The+free+encyclopedia%2E+Online+Information+Review%2C+26%286%29%3A434%2E+www%2Ewikipedia%2Eorg%2E)
10. Vladimir N. Vapnik. 1998. Statistical Learning Theory. John Wiley & Sons. [find](https://scholar.google.com/scholar?q=Vladimir+N%2E+Vapnik%2E+1998%2E+Statistical+Learning+Theory%2E+John+Wiley+%26+Sons%2E)