Publicación:
Study of a Lifelong Learning Scenario for Question Answering

dc.contributor.authorEchegoyen, Guillermo
dc.contributor.authorRodrigo Yuste, Álvaro
dc.contributor.authorPeñas Padilla, Anselmo
dc.date.accessioned2024-09-26T12:01:40Z
dc.date.available2024-09-26T12:01:40Z
dc.date.issued2022-12-15
dc.description.abstractQuestion Answering (QA) systems have witnessed a significant advance in the last years due to the development of neural architectures employing pre-trained large models like BERT. However, once the QA model is fine-tuned for a task (e.g a particular type of questions over a particular domain), system performance drops when new tasks are added along time, (e.g new types of questions or new domains). Therefore, the system requires a retraining but, since the data distribution has shifted away from the previous learning, performance over previous tasks drops significantly. Hence, we need strategies to make our systems resistant to the passage of time. Lifelong Learning (LL) aims to study how systems can take advantage of the previous learning and the knowledge acquired to maintain or improve performance over time. In this article, we explore a scenario where the same LL based QA system suffers along time several shifts in the data distribution, represented as the addition of new different QA datasets. In this setup, the following research questions arise: (i) How LL based QA systems can benefit from previously learned tasks? (ii) Is there any strategy general enough to maintain or improve the performance over time when new tasks are added? and finally, (iii) How to detect a lack of knowledge that impedes the answering of questions and must trigger a new learning process? To answer these questions, we systematically try all possible training sequencesover three well known QA datasets. Our results show how the learning of a new dataset is sensitive to previous training sequences and that we can find a strategy general enough to avoid the combinatorial explosion of testing all possible training sequences. Thus, when a new dataset is added to the system, the best way to retrain the system without dropping performance over the previous datasets is to randomly merge the new training material with the previous one.en
dc.description.versionVersión publicada VoR
dc.identifier.citationGuillermo Echegoyen, Álvaro Rodrigo, Anselmo Peñas: Study of a lifelong learning scenario for question answering. Expert Syst. Appl. 209: 118271 (2022); https://doi.org/10.1016/j.eswa.2022.118271
dc.identifier.issn0957-4174
dc.identifier.urihttps://hdl.handle.net/20.500.14468/23829
dc.journal.titleExpert Systems with Applications
dc.journal.volume209
dc.language.isoen
dc.publisherElsevier
dc.relation.centerFacultades y escuelas::E.T.S. de Ingeniería Informática
dc.relation.departmentLenguajes y Sistemas Informáticos
dc.rightsinfo:eu-repo/semantics/openAccess
dc.rights.licenseAtribución 4.0 Internacional
dc.rights.urihttps://creativecommons.org/licenses/by/4.0/
dc.subject12 Matemáticas::1203 Ciencia de los ordenadores ::1203.17 Informática
dc.subject.keywordsQuestion Answeringen
dc.subject.keywordsLifelong Learningen
dc.subject.keywordsTransfer Learningen
dc.subject.keywordsDeep Learningen
dc.titleStudy of a Lifelong Learning Scenario for Question Answeringen
dc.typeartículoes
dc.typejournal articleen
dspace.entity.typePublication
relation.isAuthorOfPublication90ababf8-3bd1-44b2-9d12-368f2c6568ac
relation.isAuthorOfPublication1e1b14bc-1284-4aef-908c-bccf31bd055e
relation.isAuthorOfPublication.latestForDiscovery90ababf8-3bd1-44b2-9d12-368f2c6568ac
Archivos
Bloque original
Mostrando 1 - 1 de 1
Cargando...
Miniatura
Nombre:
Rodrigo_Alvaro_Studyofalife.pdf
Tamaño:
330.52 KB
Formato:
Adobe Portable Document Format
Bloque de licencias
Mostrando 1 - 1 de 1
No hay miniatura disponible
Nombre:
license.txt
Tamaño:
3.62 KB
Formato:
Item-specific license agreed to upon submission
Descripción: