Leveraging Complementary AI Explanations to Mitigate Misunderstanding in XAI

Fuente: arXiv
Saved in:
Bibliographic Details
Main Authors: Xuan, Yueqing, Sokol, Kacper, Sanderson, Mark, Chan, Jeffrey
Format: Preprint
Published: 2025
Subjects:
Online Access:
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866911091211108352
author Xuan, Yueqing
Sokol, Kacper
Sanderson, Mark
Chan, Jeffrey
author_facet Xuan, Yueqing
Sokol, Kacper
Sanderson, Mark
Chan, Jeffrey
contents Artificial intelligence explanations can make complex predictive models more comprehensible. To be effective, however, they should anticipate and mitigate possible misinterpretations, e.g., arising when users infer incorrect information that is not explicitly conveyed. To this end, we propose complementary explanations -- a novel method that pairs explanations to compensate for their respective limitations. A complementary explanation adds insights that clarify potential misconceptions stemming from the primary explanation while ensuring their coherency and avoiding redundancy. We introduce a framework for designing and evaluating complementary explanation pairs based on pertinent qualitative properties and quantitative metrics. Our approach allows to construct complementary explanations that minimise the chance of their misinterpretation.
format Preprint
id arxiv_https___arxiv_org_abs_2503_00303
institution arXiv
publishDate 2025
record_format arxiv
spellingShingle Leveraging Complementary AI Explanations to Mitigate Misunderstanding in XAI
Xuan, Yueqing
Sokol, Kacper
Sanderson, Mark
Chan, Jeffrey
Human-Computer Interaction
Artificial intelligence explanations can make complex predictive models more comprehensible. To be effective, however, they should anticipate and mitigate possible misinterpretations, e.g., arising when users infer incorrect information that is not explicitly conveyed. To this end, we propose complementary explanations -- a novel method that pairs explanations to compensate for their respective limitations. A complementary explanation adds insights that clarify potential misconceptions stemming from the primary explanation while ensuring their coherency and avoiding redundancy. We introduce a framework for designing and evaluating complementary explanation pairs based on pertinent qualitative properties and quantitative metrics. Our approach allows to construct complementary explanations that minimise the chance of their misinterpretation.
title Leveraging Complementary AI Explanations to Mitigate Misunderstanding in XAI
topic Human-Computer Interaction
url https://arxiv.org/abs/2503.00303