A Design Space for Explainable Ranking and Ranking Models

Item ranking systems support users in multi-criteria decision-making tasks. Users need to trust rankings and ranking algorithms to reflect user preferences nicely while avoiding systematic errors and biases. However, today only few approaches help end users, model developers, and analysts to explain...

Full description

Saved in:
Bibliographic Details
Published inarXiv.org
Main Authors I Al Hazwani, Schmid, J, Sachdeva, M, Bernard, J
Format Paper
LanguageEnglish
Published Ithaca Cornell University Library, arXiv.org 27.05.2022
Subjects
Algorithms
Decision making
Design
End users
Multiple criterion
Ranking
Ratings & rankings
Recommender systems
Systematic errors
User groups
Online AccessGet full text

Cover

More Information
Summary:Item ranking systems support users in multi-criteria decision-making tasks. Users need to trust rankings and ranking algorithms to reflect user preferences nicely while avoiding systematic errors and biases. However, today only few approaches help end users, model developers, and analysts to explain rankings. We report on the study of explanation approaches from the perspectives of recommender systems, explainable AI, and visualization research and propose the first cross-domain design space for explainers of item rankings. In addition, we leverage the descriptive power of the design space to characterize a) existing explainers and b) three main user groups involved in ranking explanation tasks. The generative power of the design space is a means for future designers and developers to create more target-oriented solutions in this only weakly exploited space.
ISSN:2331-8422