Cross-domain polarity classification using a knowledge-enhanced meta-classifier

Current approaches to single and cross-domain polarity classification usually use bag of words, n-grams or lexical resource-based classifiers. In this paper, we propose the use of meta-learning to combine and enrich those approaches by adding also other knowledge-based features. In addition to the a...

ver descrição completa

Detalhes bibliográficos
Autores: Franco Salvador, Marc, Cruz Mata, Fermín, Troyano Jiménez, José Antonio, Rosso, Paolo
Formato: artículo
Estado:Versión enviada para evaluación y publicación
Fecha de publicación:2015
País:España
Recursos:Universidad de Sevilla (US)
Repositorio:idUS. Depósito de Investigación de la Universidad de Sevilla
OAI Identifier:oai:idus.us.es:11441/99638
Acesso em linha:https://hdl.handle.net/11441/99638
https://doi.org/10.1016/j.knosys.2015.05.020
Access Level:acceso abierto
Palavra-chave:Sentiment analysis
Cross-domain polarity classification
Meta-learning
Word sense disambiguation
Semantic network
id ES_fb8f2d53fd00072b7326d6b112dae8d1
oai_identifier_str oai:idus.us.es:11441/99638
network_acronym_str ES
network_name_str España
repository_id_str
spelling Cross-domain polarity classification using a knowledge-enhanced meta-classifierFranco Salvador, MarcCruz Mata, FermínTroyano Jiménez, José AntonioRosso, PaoloSentiment analysisCross-domain polarity classificationMeta-learningWord sense disambiguationSemantic networkCurrent approaches to single and cross-domain polarity classification usually use bag of words, n-grams or lexical resource-based classifiers. In this paper, we propose the use of meta-learning to combine and enrich those approaches by adding also other knowledge-based features. In addition to the aforementioned classical approaches, our system uses the BabelNet multilingual semantic network to generate features derived from word sense disambiguation and vocabulary expansion. Experimental results show state-of-the-art performance on single and cross-domain polarity classification. Contrary to other approaches, ours is generic. These results were obtained without any domain adaptation technique. Moreover, the use of meta-learning allows our approach to obtain the most stable results across domains. Finally, our empirical analysis provides interesting insights on the use of semantic network-based features.European Comission WIQ-EI IRSES (No. 269180)Ministerio de Economía y Competitividad TIN2012-38603-C02-01Ministerio de Economía y Competitividad TIN2012-38536-C03-02Junta de Andalucía P11-TIC-7684 MOElsevierLenguajes y Sistemas InformáticosEuropean Commission (EC)Ministerio de Economía y Competitividad (MINECO). EspañaMinisterio de Economía y Competitividad (MINECO). EspañaJunta de Andalucía2015info:eu-repo/semantics/articleinfo:eu-repo/semantics/submittedVersionapplication/pdfapplication/pdfhttps://hdl.handle.net/11441/99638https://doi.org/10.1016/j.knosys.2015.05.020reponame:idUS. Depósito de Investigación de la Universidad de Sevillainstname:Universidad de Sevilla (US)InglésKnowledge-Based Systems, 86 (september 2015), 45-56.WIQ-EI IRSES (No. 269180)TIN2012-38603-C02-01TIN2012-38536-C03-02P11-TIC-7684 MOhttps://www.sciencedirect.com/science/article/abs/pii/S0950705115002063info:eu-repo/semantics/openAccessoai:idus.us.es:11441/996382026-06-17T12:51:07Z
dc.title.none.fl_str_mv Cross-domain polarity classification using a knowledge-enhanced meta-classifier
title Cross-domain polarity classification using a knowledge-enhanced meta-classifier
spellingShingle Cross-domain polarity classification using a knowledge-enhanced meta-classifier
Franco Salvador, Marc
Sentiment analysis
Cross-domain polarity classification
Meta-learning
Word sense disambiguation
Semantic network
title_short Cross-domain polarity classification using a knowledge-enhanced meta-classifier
title_full Cross-domain polarity classification using a knowledge-enhanced meta-classifier
title_fullStr Cross-domain polarity classification using a knowledge-enhanced meta-classifier
title_full_unstemmed Cross-domain polarity classification using a knowledge-enhanced meta-classifier
title_sort Cross-domain polarity classification using a knowledge-enhanced meta-classifier
dc.creator.none.fl_str_mv Franco Salvador, Marc
Cruz Mata, Fermín
Troyano Jiménez, José Antonio
Rosso, Paolo
author Franco Salvador, Marc
author_facet Franco Salvador, Marc
Cruz Mata, Fermín
Troyano Jiménez, José Antonio
Rosso, Paolo
author_role author
author2 Cruz Mata, Fermín
Troyano Jiménez, José Antonio
Rosso, Paolo
author2_role author
author
author
dc.contributor.none.fl_str_mv Lenguajes y Sistemas Informáticos
European Commission (EC)
Ministerio de Economía y Competitividad (MINECO). España
Ministerio de Economía y Competitividad (MINECO). España
Junta de Andalucía
dc.subject.none.fl_str_mv Sentiment analysis
Cross-domain polarity classification
Meta-learning
Word sense disambiguation
Semantic network
topic Sentiment analysis
Cross-domain polarity classification
Meta-learning
Word sense disambiguation
Semantic network
description Current approaches to single and cross-domain polarity classification usually use bag of words, n-grams or lexical resource-based classifiers. In this paper, we propose the use of meta-learning to combine and enrich those approaches by adding also other knowledge-based features. In addition to the aforementioned classical approaches, our system uses the BabelNet multilingual semantic network to generate features derived from word sense disambiguation and vocabulary expansion. Experimental results show state-of-the-art performance on single and cross-domain polarity classification. Contrary to other approaches, ours is generic. These results were obtained without any domain adaptation technique. Moreover, the use of meta-learning allows our approach to obtain the most stable results across domains. Finally, our empirical analysis provides interesting insights on the use of semantic network-based features.
publishDate 2015
dc.date.none.fl_str_mv 2015
dc.type.none.fl_str_mv info:eu-repo/semantics/article
info:eu-repo/semantics/submittedVersion
format article
status_str submittedVersion
dc.identifier.none.fl_str_mv https://hdl.handle.net/11441/99638
https://doi.org/10.1016/j.knosys.2015.05.020
url https://hdl.handle.net/11441/99638
https://doi.org/10.1016/j.knosys.2015.05.020
dc.language.none.fl_str_mv Inglés
language_invalid_str_mv Inglés
dc.relation.none.fl_str_mv Knowledge-Based Systems, 86 (september 2015), 45-56.
WIQ-EI IRSES (No. 269180)
TIN2012-38603-C02-01
TIN2012-38536-C03-02
P11-TIC-7684 MO
https://www.sciencedirect.com/science/article/abs/pii/S0950705115002063
dc.rights.none.fl_str_mv info:eu-repo/semantics/openAccess
eu_rights_str_mv openAccess
dc.format.none.fl_str_mv application/pdf
application/pdf
dc.publisher.none.fl_str_mv Elsevier
publisher.none.fl_str_mv Elsevier
dc.source.none.fl_str_mv reponame:idUS. Depósito de Investigación de la Universidad de Sevilla
instname:Universidad de Sevilla (US)
instname_str Universidad de Sevilla (US)
reponame_str idUS. Depósito de Investigación de la Universidad de Sevilla
collection idUS. Depósito de Investigación de la Universidad de Sevilla
repository.name.fl_str_mv
repository.mail.fl_str_mv
_version_ 1869425340268412928
score 15,301603

Registos relacionados