ExplainabilityTransparencyAccountabilitycs.AI
Argument Attribution Explanations in Quantitative Bipolar Argumentation Frameworks (Technical Report)
Xiang Yin, Nico Potyka, Francesca Toni
July 25, 2023
European Conference on Artificial Intelligence
12 citations
Key Takeaways
-
Explainability gap in AI systems: While researchers have developed methods to explain qualitative reasoning in argumentation frameworks, explaining quantitative reasoning outcomes—which are widely used in real applications—has been largely overlooked, representing a significant gap in responsible AI practices.
-
New explanation method for AI decisions: The paper introduces Argument Attribution Explanations (AAEs), a novel approach that identifies which specific arguments influence AI system outputs, adapting techniques from machine learning to make argumentation-based AI systems more transparent and interpretable.
-
Practical applications in high-stakes domains: The researchers demonstrate AAEs' applicability through case studies in fake news detection and movie recommendations, showing that this explanation method can be implemented in real-world scenarios where understanding AI reasoning is important for user trust and accountability.
-
Establishing quality standards for explanations: The paper proposes desirable properties that good explanations should satisfy, helping establish benchmarks for what constitutes a reliable and useful explanation in argumentation-based AI systems—important for ensuring explanations are actually trustworthy rather than merely appearing transparent.
Abstract
Argumentative explainable AI has been advocated by several in recent years, with an increasing interest on explaining the reasoning outcomes of Argumentation Frameworks (AFs). While there is a considerable body of research on qualitatively explaining the reasoning outcomes of AFs with debates/disputes/dialogues in the spirit of extension-based semantics, explaining the quantitative reasoning outcomes of AFs under gradual semantics has not received much attention, despite widespread use in applications. In this paper, we contribute to filling this gap by proposing a novel theory of Argument Attribution Explanations (AAEs) by incorporating the spirit of feature attribution from machine learning in the context of Quantitative Bipolar Argumentation Frameworks (QBAFs): whereas feature attribution is used to determine the influence of features towards outputs of machine learning models, AAEs are used to determine the influence of arguments towards topic arguments of interest. We study desirable properties of AAEs, including some new ones and some partially adapted from the literature to our setting. To demonstrate the applicability of our AAEs in practice, we conclude by carrying out two case studies in the scenarios of fake news detection and movie recommender systems.