Repository logo
 
Publication

Boosting dynamic ensemble’s performance in Twitter

2019-11-09Journal article Restricted access
Simple item page
datacite.subject.fosCiências Naturais::Ciências da Computação e da Informação
dc.contributor.authorCosta, Joana
dc.contributor.authorSilva, Catarina
dc.contributor.authorAntunes, Mário
dc.contributor.authorRibeiro, Bernardete
dc.date.accessioned2025-09-16T14:25:35Z
dc.date.available2025-09-16T14:25:35Z
dc.date.issued2019-11-09
dc.description.abstractMany text classification problems in social networks, and other contexts, are also dynamic problems, where concepts drift through time, and meaningful labels are dynamic. In Twitter-based applications in particular, ensembles are often applied to problems that fit this description, for example sentiment analysis or adapting to drifting circumstances. While it can be straightforward to request different classifiers' input on such ensembles, our goal is to boost dynamic ensembles by combining performance metrics as efficiently as possible. We present a twofold performance-based framework to classify incoming tweets based on recent tweets. On the one hand, individual ensemble classifiers' performance is paramount in defining their contribution to the ensemble. On the other hand, examples are actively selected based on their ability to effectively contribute to the performance in classifying drifting concepts. The main step of the algorithm uses different performance metrics to determine both each classifier strength in the ensemble and each example importance, and hence lifetime, in the learning process. We demonstrate, on a drifted benchmark dataset, that our framework drives the classification performance considerably up for it to make a difference in a variety of applications.eng
dc.identifier.citationCosta, J., Silva, C., Antunes, M. et al. Boosting dynamic ensemble’s performance in Twitter. Neural Comput & Applic 32, 10655–10667 (2020). https://doi.org/10.1007/s00521-019-04599-7.
dc.identifier.doi10.1007/s00521-019-04599-7
dc.identifier.eissn1433-3058
dc.identifier.issn0941-0643
dc.identifier.urihttp://hdl.handle.net/10400.8/14074
dc.language.isoeng
dc.peerreviewedyes
dc.publisherSpringer Nature
dc.relation.hasversionhttps://link.springer.com/article/10.1007/s00521-019-04599-7
dc.relation.ispartofNeural Computing and Applications
dc.rights.urihttp://creativecommons.org/licenses/by/4.0/
dc.subjectDynamic ensembles
dc.subjectText classification
dc.subjectTwitter
dc.titleBoosting dynamic ensemble’s performance in Twittereng
dc.typejournal article
dspace.entity.typePublication
oaire.citation.endPage10667
oaire.citation.startPage10655
oaire.citation.titleNeural Computing and Applications
oaire.citation.volume32
oaire.versionhttp://purl.org/coar/version/c_970fb48d4fbd8a85
person.familyNameCosta
person.familyNameSilva
person.familyNameAntunes
person.givenNameJoana
person.givenNameCatarina
person.givenNameMário
person.identifierR-000-NX4
person.identifier.ciencia-id1B19-3DDC-BE75
person.identifier.ciencia-idAF10-7EDD-5153
person.identifier.orcid0000-0002-4053-5718
person.identifier.orcid0000-0002-5656-0061
person.identifier.orcid0000-0003-3448-6726
person.identifier.scopus-author-id25930820200
relation.isAuthorOfPublication23d200dc-1a81-4bd9-9a4a-0efc28af6ce4
relation.isAuthorOfPublicationee28e079-5ca7-4842-9094-372c40f75c38
relation.isAuthorOfPublicatione3e87fb0-d1d6-44c3-985d-920a5560f8c1
relation.isAuthorOfPublication.latestForDiscoveryee28e079-5ca7-4842-9094-372c40f75c38

Files

Original bundle
Now showing 1 - 1 of 1
No Thumbnail Available
Name:
Boosting dynamic ensemble’s performance in Twitter.pdf
Size:
502.18 KB
Format:
Adobe Portable Document Format
Description:
Many text classification problems in social networks, and other contexts, are also dynamic problems, where concepts drift through time, and meaningful labels are dynamic. In Twitter-based applications in particular, ensembles are often applied to problems that fit this description, for example sentiment analysis or adapting to drifting circumstances. While it can be straightforward to request different classifiers' input on such ensembles, our goal is to boost dynamic ensembles by combining performance metrics as efficiently as possible. We present a twofold performance-based framework to classify incoming tweets based on recent tweets. On the one hand, individual ensemble classifiers' performance is paramount in defining their contribution to the ensemble. On the other hand, examples are actively selected based on their ability to effectively contribute to the performance in classifying drifting concepts. The main step of the algorithm uses different performance metrics to determine both each classifier strength in the ensemble and each example importance, and hence lifetime, in the learning process. We demonstrate, on a drifted benchmark dataset, that our framework drives the classification performance considerably up for it to make a difference in a variety of applications.
Download
License bundle
Now showing 1 - 1 of 1
No Thumbnail Available
Name:
license.txt
Size:
1.32 KB
Format:
Item-specific license agreed upon to submission
Description:
Download

Collections

ESTG - Artigos em revistas internacionais