Analysis of joint inference strategies for the semantic role labeling of Spanish and Catalan

Mihai Surdeanu, Roser Morante, Lluís Màrquez

Research output: Chapter in Book/Report/Conference proceedingConference contribution

4 Citations (Scopus)

Abstract

This paper analyzes two joint inference approaches for semantic role labeling: re-ranking of candidate semantic frames generated by one local model and combination of two distinct models at argument-level using meta learning. We perform an empirical analysis on two recently released corpora of annotated semantic roles in Spanish and Catalan. This work yields several novel conclusions: (a) the proposed joint inference strategies yield good results even under adverse conditions: small training corpora, only two individual models available for combination, minimal output available from the individual models; (b) stacking of the two joint inference approaches is successful, which indicates that the two inference models provide complementary benefits. Our results are currently the best for the identification of semantic role for Spanish and Catalan.

Original languageEnglish (US)
Title of host publicationLecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
Pages206-218
Number of pages13
Volume4919 LNCS
DOIs
StatePublished - 2008
Externally publishedYes
Event9th International Conference on Computational Linguistics and Intelligent Text Processing, CICLing 2008 - Haifa, Israel
Duration: Feb 17 2008Feb 23 2008

Publication series

NameLecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
Volume4919 LNCS
ISSN (Print)03029743
ISSN (Electronic)16113349

Other

Other9th International Conference on Computational Linguistics and Intelligent Text Processing, CICLing 2008
CountryIsrael
CityHaifa
Period2/17/082/23/08

Fingerprint

Semantics
Labeling
Joints
Meta-learning
Model
Empirical Analysis
Stacking
Learning
Ranking
Strategy
Distinct
Output
Corpus

ASJC Scopus subject areas

  • Computer Science(all)
  • Biochemistry, Genetics and Molecular Biology(all)
  • Theoretical Computer Science

Cite this

Surdeanu, M., Morante, R., & Màrquez, L. (2008). Analysis of joint inference strategies for the semantic role labeling of Spanish and Catalan. In Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics) (Vol. 4919 LNCS, pp. 206-218). (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics); Vol. 4919 LNCS). https://doi.org/10.1007/978-3-540-78135-6_18

Analysis of joint inference strategies for the semantic role labeling of Spanish and Catalan. / Surdeanu, Mihai; Morante, Roser; Màrquez, Lluís.

Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics). Vol. 4919 LNCS 2008. p. 206-218 (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics); Vol. 4919 LNCS).

Research output: Chapter in Book/Report/Conference proceedingConference contribution

Surdeanu, M, Morante, R & Màrquez, L 2008, Analysis of joint inference strategies for the semantic role labeling of Spanish and Catalan. in Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics). vol. 4919 LNCS, Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics), vol. 4919 LNCS, pp. 206-218, 9th International Conference on Computational Linguistics and Intelligent Text Processing, CICLing 2008, Haifa, Israel, 2/17/08. https://doi.org/10.1007/978-3-540-78135-6_18
Surdeanu M, Morante R, Màrquez L. Analysis of joint inference strategies for the semantic role labeling of Spanish and Catalan. In Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics). Vol. 4919 LNCS. 2008. p. 206-218. (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)). https://doi.org/10.1007/978-3-540-78135-6_18
Surdeanu, Mihai ; Morante, Roser ; Màrquez, Lluís. / Analysis of joint inference strategies for the semantic role labeling of Spanish and Catalan. Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics). Vol. 4919 LNCS 2008. pp. 206-218 (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)).
@inproceedings{339cce6cb1874f05abbaaf68ede93acb,
title = "Analysis of joint inference strategies for the semantic role labeling of Spanish and Catalan",
abstract = "This paper analyzes two joint inference approaches for semantic role labeling: re-ranking of candidate semantic frames generated by one local model and combination of two distinct models at argument-level using meta learning. We perform an empirical analysis on two recently released corpora of annotated semantic roles in Spanish and Catalan. This work yields several novel conclusions: (a) the proposed joint inference strategies yield good results even under adverse conditions: small training corpora, only two individual models available for combination, minimal output available from the individual models; (b) stacking of the two joint inference approaches is successful, which indicates that the two inference models provide complementary benefits. Our results are currently the best for the identification of semantic role for Spanish and Catalan.",
author = "Mihai Surdeanu and Roser Morante and Llu{\'i}s M{\`a}rquez",
year = "2008",
doi = "10.1007/978-3-540-78135-6_18",
language = "English (US)",
isbn = "354078134X",
volume = "4919 LNCS",
series = "Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)",
pages = "206--218",
booktitle = "Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)",

}

TY - GEN

T1 - Analysis of joint inference strategies for the semantic role labeling of Spanish and Catalan

AU - Surdeanu, Mihai

AU - Morante, Roser

AU - Màrquez, Lluís

PY - 2008

Y1 - 2008

N2 - This paper analyzes two joint inference approaches for semantic role labeling: re-ranking of candidate semantic frames generated by one local model and combination of two distinct models at argument-level using meta learning. We perform an empirical analysis on two recently released corpora of annotated semantic roles in Spanish and Catalan. This work yields several novel conclusions: (a) the proposed joint inference strategies yield good results even under adverse conditions: small training corpora, only two individual models available for combination, minimal output available from the individual models; (b) stacking of the two joint inference approaches is successful, which indicates that the two inference models provide complementary benefits. Our results are currently the best for the identification of semantic role for Spanish and Catalan.

AB - This paper analyzes two joint inference approaches for semantic role labeling: re-ranking of candidate semantic frames generated by one local model and combination of two distinct models at argument-level using meta learning. We perform an empirical analysis on two recently released corpora of annotated semantic roles in Spanish and Catalan. This work yields several novel conclusions: (a) the proposed joint inference strategies yield good results even under adverse conditions: small training corpora, only two individual models available for combination, minimal output available from the individual models; (b) stacking of the two joint inference approaches is successful, which indicates that the two inference models provide complementary benefits. Our results are currently the best for the identification of semantic role for Spanish and Catalan.

UR - http://www.scopus.com/inward/record.url?scp=49949101928&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=49949101928&partnerID=8YFLogxK

U2 - 10.1007/978-3-540-78135-6_18

DO - 10.1007/978-3-540-78135-6_18

M3 - Conference contribution

AN - SCOPUS:49949101928

SN - 354078134X

SN - 9783540781349

VL - 4919 LNCS

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

SP - 206

EP - 218

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

ER -