Improving explanations: Applying the feature understandability scale for cost-sensitive feature selection

Simple item page
dc.contributor.authorRossberg, Nicola
dc.contributor.authorKleinberg, Bennett
dc.contributor.authorO'Sullivan, Barry
dc.contributor.authorLongo, Luca
dc.contributor.authorVisentin, Andrea
dc.date.accessioned2026-04-07T16:00:01Z
dc.date.available2026-04-07T16:00:01Z
dc.date.issued03/07/2026
dc.description.abstractWith the growing pervasiveness of artificial intelligence, the ability to explain the inferences made by machine learning models has become increasingly important. Numerous techniques for model explainability have been proposed, with natural-language textual explanations among the most widely used approaches. When applied to tabular data, these explanations typically draw on input features to justify a given inference. Consequently, a user’s ability to interpret the explanation depends on their understanding of the input features. To quantify this feature-level understanding, Rossberg et al. introduced the Feature Understandability Scale [49]. Building on that work, this proof-of-concept study collects understandability scores across two datasets, proposes a co-optimisation methodology of understandability and accuracy and presents the resulting explanations alongside the model accuracies. This work contributes to the body of knowledge on model interpretability by design. It is found that accuracy and understandability can be successfully co-optimised while maintaining high classification performances. The resulting explanations are considered more understandable at face value. Further research will aim to confirm these findings through user evaluation.en
dc.description.sponsorshipResearch Ireland- Taighde Éireann (Grant Nos. 18/CRT/6223; 12/RC/2289-P2)
dc.description.statusPeer reviewed
dc.description.versionAccepted Version
dc.format.extent24
dc.format.extent1110693
dc.format.mimetypeapplication/pdf
dc.identifier.authororcidRossberg, Nicola
dc.identifier.authororcidKleinberg, Bennett
dc.identifier.authororcidO'Sullivan, Barry§0000-0002-0090-2085
dc.identifier.authororcidLongo, Luca§0000-0002-2718-5426
dc.identifier.authororcidVisentin, Andrea§0000-0003-3702-4826
dc.identifier.citationRossberg, N., Kleinberg, B., O'Sullivan, B., Longo, L. and Visentin, A. (2026) 'Improving explanations: Applying the feature understandability scale for cost-sensitive feature selection', 4th World Conference on eXplainable Artificial Intelligence, Fortaleza, Brazil, 1-3 July 2026, pp. 1-24.
dc.identifier.endpage24
dc.identifier.journaltitle4th World Conference on eXplainable Artificial Intelligence
dc.identifier.otherORCID: /0000-0003-3702-4826/work/211021296
dc.identifier.otherORCID: /0000-0002-2718-5426/work/211021713
dc.identifier.startpage1
dc.identifier.urihttps://hdl.handle.net/10468/18683
dc.language.isoen
dc.rights© 2026, the authors. For the purpose of Open Access, the authors have applied a CC BY public copyright licence to any Author Accepted Manuscript version arising from this submission.
dc.subjectArtificial Intelligence and Data Analytics
dc.subjectPsychometrics
dc.subjectMachine learning
dc.subjectInterpretability by design
dc.subjectExplainable Artificial Intelligence
dc.subjectEvaluation methods
dc.subject[ComputerScience]
dc.subject[Insight Centre for Data Analytics]
dc.titleImproving explanations: Applying the feature understandability scale for cost-sensitive feature selectionen
dc.typeConference item
Files
Original bundle
Now showing 1 - 1 of 1
Thumbnail Image
Name:
Nicola_FINAL_Xai_2026.pdf
Size:
1.06 MB
Format:
Adobe Portable Document Format
Description:
Accepted Version
Download
Collections
School of Computer Science and Information Technology