Binary and multi-class classification of lexical functions in Spanish verb-noun collocations

Olga Kolesnikova; Alexander Gelbukh

doi:10.1007/978-3-030-02840-4_1

Binary and multi-class classification of lexical functions in Spanish verb-noun collocations

Olga Kolesnikova, Alexander Gelbukh

Research output: Chapter in Book/Report/Conference proceeding › Conference contribution › peer-review

1 Scopus citations

Abstract

Collocations as semi-fixed lexical combinations present a challenge in natural language processing. While collocation identification on the shallow level is a task in which a significant advance has been reached, a deeper semantic representation and analysis of collocations remains an open issue. One of the possible solutions is detection of lexical functions of the Meaning-Text Theory in collocations thus resolving their semantic interpretation. We experimented with four lexical functions (Oper1, Real1, CausFunc0, and CausFunc1) for the special case of Spanish verb-noun collocations. In our experiments we also identified free verb-noun combinations as opposed to lexical functions. We used WordNet hypernyms as features and various algorithms of supervised machine learning; the best result with an F-measure of 0.873 was achieved for detecting Oper1 in binary classification.

Original language	English
Title of host publication	Advances in Computational Intelligence - 16th Mexican International Conference on Artificial Intelligence, MICAI 2017, Proceedings
Editors	Miguel González-Mendoza, Félix Castro, Sabino Miranda-Jiménez
Publisher	Springer Verlag
Pages	3-14
Number of pages	12
ISBN (Print)	9783030028398
DOIs	https://doi.org/10.1007/978-3-030-02840-4_1
State	Published - 2018
Event	16th Mexican International Conference on Artificial Intelligence, MICAI 2017 - Enseneda, Mexico Duration: 23 Oct 2017 → 28 Oct 2017

Publication series

Name	Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
Volume	10633 LNAI
ISSN (Print)	0302-9743
ISSN (Electronic)	1611-3349

Conference

Conference	16th Mexican International Conference on Artificial Intelligence, MICAI 2017
Country/Territory	Mexico
City	Enseneda
Period	23/10/17 → 28/10/17

Keywords

Hypernyms
Lexical functions
Spanish verb-noun collocations
Supervised learning

Access to Document

10.1007/978-3-030-02840-4_1

Cite this

Kolesnikova, O., & Gelbukh, A. (2018). Binary and multi-class classification of lexical functions in Spanish verb-noun collocations. In M. González-Mendoza, F. Castro, & S. Miranda-Jiménez (Eds.), Advances in Computational Intelligence - 16th Mexican International Conference on Artificial Intelligence, MICAI 2017, Proceedings (pp. 3-14). (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics); Vol. 10633 LNAI). Springer Verlag. https://doi.org/10.1007/978-3-030-02840-4_1

Kolesnikova, Olga ; Gelbukh, Alexander. / Binary and multi-class classification of lexical functions in Spanish verb-noun collocations. Advances in Computational Intelligence - 16th Mexican International Conference on Artificial Intelligence, MICAI 2017, Proceedings. editor / Miguel González-Mendoza ; Félix Castro ; Sabino Miranda-Jiménez. Springer Verlag, 2018. pp. 3-14 (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)).

@inproceedings{df4dbafed9984c518c7ecbb1f5f8c199,

title = "Binary and multi-class classification of lexical functions in Spanish verb-noun collocations",

abstract = "Collocations as semi-fixed lexical combinations present a challenge in natural language processing. While collocation identification on the shallow level is a task in which a significant advance has been reached, a deeper semantic representation and analysis of collocations remains an open issue. One of the possible solutions is detection of lexical functions of the Meaning-Text Theory in collocations thus resolving their semantic interpretation. We experimented with four lexical functions (Oper1, Real1, CausFunc0, and CausFunc1) for the special case of Spanish verb-noun collocations. In our experiments we also identified free verb-noun combinations as opposed to lexical functions. We used WordNet hypernyms as features and various algorithms of supervised machine learning; the best result with an F-measure of 0.873 was achieved for detecting Oper1 in binary classification.",

keywords = "Hypernyms, Lexical functions, Spanish verb-noun collocations, Supervised learning",

author = "Olga Kolesnikova and Alexander Gelbukh",

note = "Publisher Copyright: {\textcopyright} Springer Nature Switzerland AG 2018.; 16th Mexican International Conference on Artificial Intelligence, MICAI 2017 ; Conference date: 23-10-2017 Through 28-10-2017",

year = "2018",

doi = "10.1007/978-3-030-02840-4_1",

language = "Ingl{\'e}s",

isbn = "9783030028398",

series = "Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)",

publisher = "Springer Verlag",

pages = "3--14",

editor = "Miguel Gonz{\'a}lez-Mendoza and F{\'e}lix Castro and Sabino Miranda-Jim{\'e}nez",

booktitle = "Advances in Computational Intelligence - 16th Mexican International Conference on Artificial Intelligence, MICAI 2017, Proceedings",

address = "Alemania",

}

Kolesnikova, O & Gelbukh, A 2018, Binary and multi-class classification of lexical functions in Spanish verb-noun collocations. in M González-Mendoza, F Castro & S Miranda-Jiménez (eds), Advances in Computational Intelligence - 16th Mexican International Conference on Artificial Intelligence, MICAI 2017, Proceedings. Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics), vol. 10633 LNAI, Springer Verlag, pp. 3-14, 16th Mexican International Conference on Artificial Intelligence, MICAI 2017, Enseneda, Mexico, 23/10/17. https://doi.org/10.1007/978-3-030-02840-4_1

Binary and multi-class classification of lexical functions in Spanish verb-noun collocations. / Kolesnikova, Olga ; Gelbukh, Alexander.
Advances in Computational Intelligence - 16th Mexican International Conference on Artificial Intelligence, MICAI 2017, Proceedings. ed. / Miguel González-Mendoza; Félix Castro; Sabino Miranda-Jiménez. Springer Verlag, 2018. p. 3-14 (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics); Vol. 10633 LNAI).

Research output: Chapter in Book/Report/Conference proceeding › Conference contribution › peer-review

TY - GEN

T1 - Binary and multi-class classification of lexical functions in Spanish verb-noun collocations

AU - Kolesnikova, Olga

AU - Gelbukh, Alexander

N1 - Publisher Copyright: © Springer Nature Switzerland AG 2018.

PY - 2018

Y1 - 2018

N2 - Collocations as semi-fixed lexical combinations present a challenge in natural language processing. While collocation identification on the shallow level is a task in which a significant advance has been reached, a deeper semantic representation and analysis of collocations remains an open issue. One of the possible solutions is detection of lexical functions of the Meaning-Text Theory in collocations thus resolving their semantic interpretation. We experimented with four lexical functions (Oper1, Real1, CausFunc0, and CausFunc1) for the special case of Spanish verb-noun collocations. In our experiments we also identified free verb-noun combinations as opposed to lexical functions. We used WordNet hypernyms as features and various algorithms of supervised machine learning; the best result with an F-measure of 0.873 was achieved for detecting Oper1 in binary classification.

AB - Collocations as semi-fixed lexical combinations present a challenge in natural language processing. While collocation identification on the shallow level is a task in which a significant advance has been reached, a deeper semantic representation and analysis of collocations remains an open issue. One of the possible solutions is detection of lexical functions of the Meaning-Text Theory in collocations thus resolving their semantic interpretation. We experimented with four lexical functions (Oper1, Real1, CausFunc0, and CausFunc1) for the special case of Spanish verb-noun collocations. In our experiments we also identified free verb-noun combinations as opposed to lexical functions. We used WordNet hypernyms as features and various algorithms of supervised machine learning; the best result with an F-measure of 0.873 was achieved for detecting Oper1 in binary classification.

KW - Hypernyms

KW - Lexical functions

KW - Spanish verb-noun collocations

KW - Supervised learning

UR - http://www.scopus.com/inward/record.url?scp=85059933472&partnerID=8YFLogxK

U2 - 10.1007/978-3-030-02840-4_1

DO - 10.1007/978-3-030-02840-4_1

M3 - Contribución a la conferencia

SN - 9783030028398

T3 - Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)

SP - 3

EP - 14

BT - Advances in Computational Intelligence - 16th Mexican International Conference on Artificial Intelligence, MICAI 2017, Proceedings

A2 - González-Mendoza, Miguel

A2 - Castro, Félix

A2 - Miranda-Jiménez, Sabino

PB - Springer Verlag

T2 - 16th Mexican International Conference on Artificial Intelligence, MICAI 2017

Y2 - 23 October 2017 through 28 October 2017

ER -

Kolesnikova O , Gelbukh A. Binary and multi-class classification of lexical functions in Spanish verb-noun collocations. In González-Mendoza M, Castro F, Miranda-Jiménez S, editors, Advances in Computational Intelligence - 16th Mexican International Conference on Artificial Intelligence, MICAI 2017, Proceedings. Springer Verlag. 2018. p. 3-14. (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)). doi: 10.1007/978-3-030-02840-4_1

Binary and multi-class classification of lexical functions in Spanish verb-noun collocations

Abstract

Publication series

Conference

Keywords

Access to Document

Other files and links

Fingerprint

Cite this