Abstract
The feature selection method we are presenting in this chapter makes use of syntactic knowledge provided by dependency relations. Dependency-based feature selection for the Naïve Bayes model is examined and exemplified in the case of adjectives. Performing this type of knowledge-based feature selection places the disambiguation process at the border between unsupervised and knowledge-based techniques. The discussed type of feature selection and corresponding disambiguation method will once again prove that a basic, simple knowledge-lean disambiguation algorithm, hereby represented by the Naïve Bayes model, can perform quite well when provided knowledge in an appropriate way. Our main conclusion will be that the Naïve Bayes model reacts well in the presence of syntactic knowledge of this type and that dependency-based feature selection for the Naïve Bayes model is a reliable alternative to the WordNet-based semantic one.
This is a preview of subscription content, log in via an institution to check access.
Access this chapter
Tax calculation will be finalised at checkout
Purchases are for personal use only
Notes
- 1.
Dependency grammar (DG) is a class of syntactic theories developed by Lucien Tesnière (1959). Within this theory, syntactic structure is determined by the grammatical relations existing between a word (a head) and its dependents.
- 2.
See the mathematical model presented in Chap. 2.
- 3.
The relations between the dependent and the head are usually represented by an arch.
- 4.
- 5.
For which see Sect. 4.3.1.
- 6.
See the mathematical model presented in Chap. 2.
- 7.
They have only eliminated the potentially unuseful relations—for WSD—provided by the Stanford parser, such as: determiner, predeterminer, numeric determiner, punctuation relations, etc.
- 8.
In what follows, such dependencies will be called first order dependencies.
- 9.
A path anchored at the target word w is a path in the dependency graph starting at w. If the dependency relations have directionality, leading to an associated oriented graph, a path anchored at w is either a path starting at w or arriving at w.
- 10.
In what follows, such dependencies will be called second order dependencies.
- 11.
Which are the same as those showing the distribution of senses of common and public, respectively in Chap. 3.
- 12.
Undertaken from (Hristea and Colhon 2012).
- 13.
Undertaken from (Hristea and Colhon 2012).
- 14.
The considered directionality is from head to dependent.
- 15.
In what follows, these dependencies will be called head-driven dependencies.
- 16.
In what follows, these dependencies will be called dependent-driven dependencies.
- 17.
The case Two head-driven dependencies can be summarized as follows: let us denote the target word by \(A\); collect all words of type \(B\) and \(C\) such that \(B\) is a dependent of \(A\) and \(C\) is a dependent of \(B.\)
- 18.
The case Head-driven dependencies and dependent-driven dependencies can be summarized as follows: let us denote the target word by \(A\); collect all words of type \(B\) and \(C\) such that \(B\) is a dependent of \(A\) and \(B\) is a dependent of \(C.\)
- 19.
The case Two dependent-driven dependencies can be summarized as follows: let us denote the target word by \(A\); collect all words of type \(B\) and \(C\) such that \(A\) is a dependent of \(B\) and \(B\) is a dependent of \(C.\)
- 20.
The case Dependent-driven dependencies and head-driven dependencies can be summarized as follows: let us denote the target word by \(A\); collect all words of type \(B\) and \(C\) such that \(A\) is a dependent of \(B\) and \(C\) is a dependent of \(B.\)
- 21.
This principle, which gives the nominal information priority, while the adjectival information is evaluated strictly within the range allowed by the nominal one, has guided Hristea and Colhon (2012) when choosing the nominal subject relation, for instance. This relation refers to the predicative form of the adjective linked via a copula verb to the noun that the adjective modifies.
- 22.
See the mathematical model presented in Chap. 2.
- 23.
Undertaken from (Hristea and Colhon 2012).
- 24.
Undertaken from (Hristea and Colhon 2012).
- 25.
This is the approach suggested by the first series of performed experiments, which had disregarded the dependency type. Test results have shown (see Sect. 4.3.2) that directionality of the relations counts and that the best disambiguation results are obtained when the target word plays the role of head.
- 26.
- 27.
Undertaken from (Hristea and Colhon 2012).
- 28.
See the mathematical model presented in Chap. 2.
- 29.
Let us note that accuracy is always higher in the case Two head-driven dependencies than in the case Head-driven dependencies and dependent-driven dependencies, which shows that, in the case of directed first and second order dependencies, it is essential to consider the head role not only of the target word but also of its dependents.
- 30.
See the mathematical model presented in Chap. 2.
- 31.
- 32.
Which allows the arches denoting the dependency relations to intersect.
- 33.
Which does not allow the arches denoting the dependency relations to intersect, in accordance with the classical dependency linguistic theory.
- 34.
See the mathematical model presented in Chap. 2.
- 35.
References
Bruce, R., Wiebe, J., Pedersen, T.: The Measure of a Model, CoRR, cmp-lg/9604018 (1996)
Chen, P., Bowes, C., Ding, W., Brown, D.: A fully unsupervised word sense disambiguation method using dependency knowledge. In: Human Language Technologies: The 2009 Annual Conference of the North American Chapter of the ACL, pp. 28–36 (2009)
de Marneffe, M.C., Manning, C.D.: Stanford typed dependencies manual. Technical Report, Stanford University (2008)
de Marneffe, M.C., MacCartney, B., Manning, C.D.: Generating typed dependency parses from phrase structure parses. In: Proceedings of LREC-06, pp. 449–454 (2006)
Grefenstette, G.: Explorations in Automatic Thesaurus Discovery. Kluwer Academic Publishers, Dordrecht (1994)
Hristea, F.: Recent advances concerning the usage of the Naïve Bayes model in unsupervised word sense disambiguation. Int. Rev. Comput. Softw. 4(1), 58–67 (2009)
Hristea, F., Colhon, M.: Feeding syntactic versus semantic knowledge to a knowledge-lean unsupervised word sense disambiguation algorithm with an underlying Naïve Bayes model. Fundam. Inform. 119(1), 61–86 (2012)
Hristea, F., Popescu, M.: Adjective sense disambiguation at the border between unsupervised and knowledge-based techniques. Fundam. Inform. 91(3–4), 547–562 (2009)
Hristea, F., Popescu, M., Dumitrescu, M.: Performing word sense disambiguation at the border between unsupervised and knowledge-based techniques. Artif. Intell. Rev. 30(1), 67–86 (2008)
Hudson, R.A.: Word Grammar. Blackwell, Oxford (1984)
Klein, D., Manning, C.D.: Accurate unlexicalized parsing. In: Proceedings of the 41st Meeting of the Association for Computational Linguistics (ACL 2003), pp. 423–430 (2003)
Lee, L.: Measures of distributional similarity. In: Proceedings of the 37th Annual Meeting of the Association for Computational Linguistics, pp. 25–32 (1999)
Levin, B.: English Verb Classes and Alternations: A Preliminary Investigation. University of Chicago Press, Chicago (1993)
Lin, D.: Automatic retrieval and clustering of similar words. In: Proceedings of the Joint Annual Meeting of the Association for Computational Linguistics and International Conference on Computational Linguistics, pp. 768–774 (1998)
Năstase, V.: Unsupervised all-words word sense disambiguation with grammatical dependencies. In: Proceedings of the Third International Joint Conference on Natural Language Processing, pp. 757–762 (2008)
Padó, S., Lapata, M.: Dependency-based construction of semantic space models. Comput. Linguist. 33(2), 161–199 (2007)
Ponzetto, S.P., Navigli, R.: Knowledge-rich word sense disambiguation rivaling supervised systems. In: Proceedings of the 48th Annual Meeting of the Association for Computational Linguistics (ACL 2010), Uppsala, Sweden, ACL Press, pp. 1522–1531 (2010)
Sleator, D., Temperley, D.: Parsing English with a link grammar. Technical Report CMU-CS-91-196, Carnegie Mellon University, Pittsburgh, PA (1991)
Sleator, D., Temperley, D.: Parsing English with a link grammar. In: Proceedings of the Third International Workshop on Parsing Technologies (IWPT93), pp. 277–292 (1993)
Tesnière, L.: Eléments de syntaxe structurale. Klincksieck, Paris (1959)
Author information
Authors and Affiliations
Corresponding author
Rights and permissions
Copyright information
© 2013 The Author(s)
About this chapter
Cite this chapter
Hristea, F.T. (2013). Syntactic Dependency-Based Feature Selection. In: The Naïve Bayes Model for Unsupervised Word Sense Disambiguation. SpringerBriefs in Statistics. Springer, Berlin, Heidelberg. https://doi.org/10.1007/978-3-642-33693-5_4
Download citation
DOI: https://doi.org/10.1007/978-3-642-33693-5_4
Published:
Publisher Name: Springer, Berlin, Heidelberg
Print ISBN: 978-3-642-33692-8
Online ISBN: 978-3-642-33693-5
eBook Packages: Mathematics and StatisticsMathematics and Statistics (R0)