Publication:

Aligning AI Explanations with User Needs: A Qualitative Study of XAI Methods

 
cris.virtual.department#PLACEHOLDER_PARENT_METADATA_VALUE#
cris.virtual.orcid#PLACEHOLDER_PARENT_METADATA_VALUE#
cris.virtualsource.departmentb3b498f5-e353-476d-8261-56bdfbbd8fbe
cris.virtualsource.orcidb3b498f5-e353-476d-8261-56bdfbbd8fbe
dc.contributor.authorDonoso-Guzman, Ivania
dc.contributor.authorParra, Denis
dc.contributor.authorVerbert, Katrien
dc.contributor.orcidext0000-0001-6699-7710
dc.date.accessioned2026-11-05T14:28:51Z
dc.date.available2026-11-05T14:28:51Z
dc.date.createdwos2026
dc.date.issued2026
dc.description.abstractIn recent years, significant progress has been made in Natural Language Processing (NLP), yet a gap persists in tailoring explanations for text classification to non-AI users. In this research, we conduct a qualitative study to compare the suitability of several explanation alternatives generated with existing Explainable AI (XAI) methods using visualisations and textual explanations created by LLMs. Our preliminary findings reveal that explanations emphasising the paper’s methodology enhance user confidence: while textual explanations are generally preferred, effective visualisations highlighting methodological aspects are valuable. Finally, results show that what users expect from explanations does not align with current efforts in XAI: they need explanations that support how they conduct the same task and are not concerned with fidelity to the model’s inner workings.
dc.description.wosFundingTextThis work was partially funded by ANID Chile, Millennium Science Initiative Program, codes ICN2021_004 (iHealth) and ICN17_002 (IMFD), by Basal Funds for Center of Excellence FB210017 (CENIA), the Research Foundation Flanders (FWO, grant GOA4923N) and KU Leuven (grant C14/21/072). In addition, we thank Fondecyt grant 1231724. The research of Ivania Donoso-Guzman was supported by the doctoral scholarship of ANID BECAS/DOCTORADO NACIONAL 21202228.
dc.identifier.doi10.1007/978-3-032-19105-2_10
dc.identifier.eissn1865-0937
dc.identifier.isbn978-3-032-19104-5
dc.identifier.issn1865-0929
dc.identifier.urihttps://imec-publications.be/handle/20.500.12860/60745
dc.language.isoeng
dc.provenance.editstepusergreet.vanhoof@imec.be
dc.publisherSPRINGER INTERNATIONAL PUBLISHING AG
dc.relation.ispartofseriesCommunications in Computer and Information Science
dc.source.beginpage140
dc.source.conferenceMachine Learning and Principles and Practice of Knowledge Discovery in Databases International Workshops of ECML PKDD
dc.source.conferencedate2025-09-19
dc.source.conferencelocationPorto
dc.source.endpage150
dc.source.journalMACHINE LEARNING AND PRINCIPLES AND PRACTICE OF KNOWLEDGE DISCOVERY IN DATABASES, ECML PKDD 2025, PT IV
dc.source.numberofpages11
dc.title

Aligning AI Explanations with User Needs: A Qualitative Study of XAI Methods

dc.typeProceedings paper
dspace.entity.typePublication
imec.internal.crawledAt2026-07-14
imec.internal.sourcecrawler
imec.internal.wosCreatedAt2026-09-07
Files
Publication available in collections: