Elamur Research⌕
☰
home / papers / 2016 / Dictionaries as Networks: Identifying the graph structure of Ogden's Basic English

Dictionaries as Networks: Identifying the graph structure of Ogden's Basic English

Paper · 2016 · Camilo Garrido and Claudio Gutierrez

Summary

The paper asks whether treating a dictionary as a graph, where each headword points to the words used in its definition, can reveal linguistically meaningful structure. As a test case, the authors check whether purely graph-theoretic tools can recover Ogden's Basic English, an 850-word list that C. K. Ogden designed as a minimal language.

They build a deliberately simple network from the open OPTED dictionary. Parts of speech are merged, each definition is reduced to an unordered set of words, inflected forms are dropped, and prepositions and articles are removed. They then compare classic social-network centralities (degree, PageRank, closeness, betweenness) against frequency-based and random baselines. They also apply several connectivity filters.

Centrality alone does not pick out Ogden's list. Its higher-ranked part resembles central words, while its weaker part looks almost random, which suggests that part of the list exists to cover the graph. Connectivity cores work much better: a mutually-defining core of about 9.3k words keeps roughly 97% of Ogden's words. The work argues that dictionary graphs need their own analysis tools rather than ones borrowed from social networks.

Key points

  • Data. OPTED (Online Plain Text English Dictionary) was chosen because it is open and reproducible and contains 94.5% of Ogden's words. That gives 803 Ogden words in the network.
  • SCC. The strongly connected component of the OPTED graph has 23,360 nodes and keeps 802 of the 803 Ogden words. About three quarters of the dictionary is discarded as terminal or iteratively terminal words.
  • BC. The Bidirectional Component links two words only if each appears in the other's definition (e.g. fire/light). It has 16,750 words and 790 Ogden words.
  • BSCC. The largest strongly connected piece of BC has 9,344 nodes and 784 Ogden words, so only about 3% of Ogden is lost. Its size is close to the roughly 10% "grounding kernel" reported by Picard et al. (2009).
  • Group centrality. Ogden was split into a top third and a bottom third (268 words each) and compared with equal-sized frequency-based and random sets.
    • The top third lines up with classic centrality.
    • The bottom third behaves like the random set on PageRank and has a much lower degree.
    • Exact group centrality could not be computed, because the problem is NP-complete.

Interesting ideas

  • I1. Turning definitions into word→word edges and keeping only words that mutually define each other gives a compact, self-referential "defining vocabulary" with no corpus statistics at all.
  • I2. Repeatedly stripping words that define nothing else is a cheap, transparent way to separate foundational vocabulary from leaf vocabulary.
  • I3. A curated word set is judged against same-size frequency and random sets, so a striking number is only meaningful relative to chance.
  • I4. Splitting a hand-made list into its high-ranked and low-ranked halves showed two roles: hub words and words that spread across the graph to cover it.
  • I5. The authors start from the most naive untyped graph and plan to add part-of-speech and sense labels later, so each refinement's contribution stays measurable.

For Mova

How these ideas could serve Mova, our small open language model — its files, stages and tests:

  • I1 (world): A gloss graph over WordNet could yield a small primitive core, used as the explainable base layer of the ontology onto which other concepts are reduced.
  • I2 (world): The same pruning could rank compiled knowledge entries by how foundational they are, deciding what must be loaded first and what can be lazily attached.
  • I3 (learn): Any induced lexicon or rule set should be reported next to random and frequency-matched controls of equal size, as a standard negative-control check.
  • I4 (back): A simplified-vocabulary generator or compressor could choose target words for coverage of the meaning graph, not only for frequency or centrality.
  • I5 (world): A knowledge graph could be built the same way, as a plain baseline first, with every added edge type tested for whether it actually changes results.

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 ChileSantiago · Chile3 works · 2,017 citations

 Fields, topics and concepts

📚 Shelves 5The field of science this work belongs to (OpenAlex fields).Arts and Humanities · 17,959 worksLanguage and Linguistics · 17,789 worksLiterature and Literary Theory · 542 worksPhysics and Astronomy · 70 worksStatistical and Nonlinear Physics · 60 works

🧭 Topics 1Research subjects and languages the work is about.English · 2,381 works

Cite

Cite as Camilo Garrido and Claudio Gutierrez. 2016. Dictionaries as Networks: Identifying the graph structure of Ogden's Basic English. Proceedings of COLING 2016, the 26th International Conference on Computational Linguistics: Technical Papers.

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

@inproceedings{garrido2016dictionaries,
  title = {{Dictionaries as Networks: Identifying the graph structure of Ogden's Basic English}},
  author = {Camilo Garrido and Claudio Gutierrez},
  booktitle = {Proceedings of COLING 2016, the 26th International Conference on Computational Linguistics: Technical Papers},
  year = {2016},
  url = {https://aclanthology.org/C16-1336/}
}

References

33 references, 3 of them with a card of ours. ↗ leads to the work, «find» searches for one we have no id for yet.

  1. Joshua T. Abbott, Joseph T. Austerweil, Thomas L. Griffiths. 2012. Human memory search as a random walk in a semantic Network. Advances in Neural Information Processing Systems. 25, 3050-58, 2012. find
  2. Robert Alfred Amsler. 1980. The Structure of the Merriam-Webster Pocket Dictionary. PhD. Thesis. Department of Computer Sciences. University of Texas. Austin, Texas. find
  3. Vladimir Bagatelj, Andrej Mrvar and Matjaž Zaveršnik. 2002. Network analysis of dictionaries. University of Ljubljana, Institute of Mathematics, Physics and Mechanics. Dept. of Theoretical Computer Science. Preprint series, Vol. 40 (2002), 834. find
  4. Nicoletta Calzolari. 1984. Detecting Patterns in a Lexical Data Base. Proceedings of the 10th International Conference on Computational Linguistics and 22nd annual meeting on Association for Computational Linguistics. Association for Computational Linguistics, 1984. find
  5. Nicoletta Calzolari and Eugenio Picchi. 1988. Acquisition of Semantic Information From an On-Line Dictionary. Proceedings of the 12th conference on Computational linguistics-Volume 1. Association for Computational Linguistics, 1988. find
  6. Elizabeth Costenbader and Thomas W. Valente. 2003. The stability of centrality measures when networks are sampled. Social Networks. 25 (2003) 283-307 find
  7. Peter Csermely, András London, Ling-Yun Wu, and Brian Uzzi 2013. Structure and dynamics of core/periphery networks. Journal of Complex Networks 1.2 (2013): 93-123. find
  8. William Dolan, Lucy Vanderwende, and Stephen D. Richardson. Automatically deriving structured knowledge bases from on-line dictionaries. Proceedings of the First Conference of the Pacific Association for Computational Linguistics. 1993. find
  9. S. N. Dorogovtsev, J. F. F. Mendes. 2001. Language as an Evolving Word Web Proceedings of the Royal Society of London B: Biological Sciences. 2001, 268 2603-2606. find
  10. M. G. Everett and S. P. Borgatti. 1999. The Centrality of Groups and Classes. Journal of Mathematical Sociology, 23(3): 181-201. find
  11. Linton C. Freeman. 1978. Centrality in Social Networks. Conceptual Clarification. Social Networks. 1 (1978/79) 215-239. find
  12. Camilo Garrido, Ricardo Mora and Claudio Gutierrez. 2016. Group Centrality for Semantic Networks: a SWOT analysis featuring Random Walks. ArXiv Preprints, https://arxiv.org/abs/1601.00139 ↗
  13. Thad Hughes, Daniel Ramage. 2007. Lexical semantic relatedness with random graph walks. EMNLP-CoNLL, 581-589. find
  14. Jure Leskovec and Andrej Krevl. 2014. SNAP Datasets: Stanford Large Network Dataset Collection. http://snap.stanford.edu/data David Levary, Jean-Pierre Eckmann, Elisha Moses, and Tsvi Tlusty. 2012. Loops and Self-Reference in the Construction of Dictionaries. Physical Review X, 2, 031018 (2012). ↗
  15. Ken C. Litkowski. 1978. Models of the Semantic Structure of Dictionaries. American Journal of Computational Linguistics, Microfiche 81, Frames 25-74. find
  16. Christopher D. Manning, Mihai Surdeanu, John Bauer, Jenny Finkel, Steven J. Bethard and David McClosky. 2014. The Stanford CoreNLP Natural Language Processing Toolkit Proceedings of the 52nd Annual Meeting of the Association for Computational Linguistics: System Demonstrations, pp. 55-60.
  17. John Matta, Jeffrey Borwey and Gunes Ercal. 2014. Comparative Resilience Notions and Vertex Attack Tolerance of Scale-Free Networks. ArXiv Preprints, http://arxiv.org/abs/1404.0103 ↗
  18. Rada Mihalcea, Paul Tarau and Elizabeth Figa. 2004. PageRank on Semantic Networks, with Application to Word Sense Disambiguation. Proc. COLING ’04.
  19. Rada Mihalcea and Paul Tarau. 2004. TextRank Bringing Order into Texts Proc. Association for Computational Linguistics 2004. find
  20. George A. Miller. 1995. WordNet: A lexical database for English. Communications of the ACM, 38(11):39-41. ↗
  21. Roberto Navigli and Simone Paolo Ponzetto. 2012. BabelNet: The Automatic Construction, Evaluation and Application of a Wide-Coverage Multilingual Semantic Network. Artificial Intelligence, 193:217-250. ↗
  22. Mark Newman. 2003. The structure and function of complex networks. SIAM review, 45(2):167-256. find
  23. Charles Kay Ogden. 1930. Basic English: A General Introduction with Rules and Grammar, London, K. Paul, Trench, Trubner & Co. find
  24. Olivier Picard, Alexandre Blondin Massé, Stevan Harnard, Odile Marcotte, Guillaume Chicoisne, Yassine Gargouri. 2009. Hierarchies in Dictionary Definition Space. 23rd. Annual Conf. on Neural Information Proc. Systems: Workshop on Analyzing Networks and Learning With Graphs, Vancouver BC, Canada, 11-12 Dec. 2009. Preprint arXiv:0911.5703. ↗
  25. Alain Polguère. 2014. From Writing Dictionaries to Weaving Lexical Networks. find
  26. Int J Lexicography, (2014) 27 (4): 396-418. find
  27. Mark Steyvers and Joshua B. Tenenbaum. 2005. The Large-scale structure of semantic networks: Statistical analyses and a model of semantic growth. Cognitive Science, 29 (1): 41-78. find
  28. Stanley Wasserman and Katherine Faust. 1994. Social Network Analysis. Methods and Applications, Cambridge Univ. Press, Cambridge, UK. find
  29. Yorick Wilks 1993. Providing machine tractable dictionary tools. Semantics and the Lexicon. find
  30. Springer Netherlands, 1993. 341-401. find
  31. Yorick Wilks, Dan Fass, Cheng-ming Guo, James E. McDonald, Tony Plate, and Brian M. Slator. 1988. Machine Tractable Dictionaries as Tools and Resources for Natural Language Processing. Proceedings of the 12th conference on Computational linguistics-Volume 2. Association for Computational Linguistics. find
  32. Eric Yeh, Daniel Ramage, Christopher D. Manning, Eneko Agirre, Aitor Soroa. 2009. Wikiwalk: random walks on wikipedia for semantic relatedness. Workshop on Graph-based Methods for NLP, 4149, Stroudsburg, PA, USA. find
  33. Nicola Zingarelli. 1970 Vocabolario della lingua italiana, a cura di Miro Dogliotti, Luigi Rosiello & Paolo Valesio. Bologna: Zanichelli (1970). find

Updated

↑ back to top

This page in Markdown → https://md.research.elamur.ai/free/en/papers/garrido-2016-dictionaries-as-networks-identifying-graph/