Argumentation and Causal Models in Human-Machine Interaction: A Round Trip

Yann Munro (1, 2) , Isabelle Bloch (2) , Mohamed Chetouani (1, 3) , Marie-Jeanne Lesot (2) , Catherine Pelachaud (1, 3)

1 ISIR - Institut des Systèmes Intelligents et de Robotique
2 LFI - Learning, Fuzzy and Intelligent systems
3 PIROS - Perception, Interaction, Robotique sociales

In the field of explainable artificial intelligence (XAI), causal models and abstract argumentation frameworks are two formal approaches that provide a definition of an explanation. These symbolic approaches rely on logical formalisms to reason by abduction or to search for causalities, from the formal modeling of a problem or a situation. These models are designed to satisfy a number of properties found in explanations given by one human to another which are particularly interesting for humanmachine interactions as well. In this paper, we show the equivalence between a particular type of causal models, that we call argumentative causal graphs (ACG), and abstract argumentation frameworks. We also propose a transformation between these two systems and look at how one definition of an explanation in the argumentation theory is transposed when moving to ACG. To illustrate our proposition, we use a very simplified version of a screening agent for COVID-19.

Mots clés: Causal models; Abstract argumentation frameworks; Explainable artificial intelligence (XAI)

Domaines: Intelligence artificielle [cs.AI]

Yann Munro, Isabelle Bloch, Mohamed Chetouani, Marie-Jeanne Lesot, Catherine Pelachaud. Argumentation and Causal Models in Human-Machine Interaction: A Round Trip. 8th International Workshop on Artificial Intelligence and Cognition, Jun 2022, Örebro, Sweden. ⟨hal-03739310⟩