SSA 2024

Abstract: In this tutorial we show how logical argumentation provides a natural model of defeasible reasoning. We reason defeasibly whenever we infer conclusions that may be retracted in exceptional, conflicting circumstances. Nonmonotonic logics model such reasoning. It is the ambition of formal argumentation to provide a unifying framework for such logics. In this tutorial, we explain some of the core methods of logical argumentation by adopting a proof-theoretic perspective in which arguments are premise-conclusions pairs (so-alled sequents) generated by a sequent-proof calculus. We discuss how conflicts between arguments are tracked by means of various types of argumentative attacks. This gives rise to a so-called argumentation framework from which arguments can be selected following the rationale of an argumentation semantics, as defined by Dung in his seminal work on abstract argumentation. We illustrate this method by focusing on reasoning with defaults in their deontic interpretation (as norms generating obligations) and their doxastic interpretation (as applying to and generating candidates for beliefs) and discuss important meta-theoretic desiderata for argumentative models of defeasible reasoning.

Abstract: Cognitive Argumentation is a computational framework for dialectic argumentation-based reasoning, built from a theoretical framework of argumentation in AI and grounded via cognitive principles from Cognitive Science. This tutorial will cover three typical reasoning tasks: The suppression task, the selection task and the syllogistic reasoning task. Some participants' responses in these psychological experiments deviate from classical logic. We will identify these reasoning patterns and formalize them as cognitive principles in Cognitive Argumentation. Finally, we will demonstrate how cognitive argumentation can be integrated into ACT-R, a cognitive architecture, where the argumentation process is guided by the context through the spreading activation of chunks, bridging to lower levels of cognition.

Abstract: Approximation Fixpoint Theory is an algebraic framework for studying semantics of knowledge representation formalisms. It builds upon the idea that knowledge bases can be associated with operators such that fixpoints of the operator correspond to models of the knowledge base. This is an established technique in logic programming, and we will show how it can be fruitfully applied to (abstract) argumentation.

Abstract: While computational argumentation has established itself as an important subfield of AI, actual applications are few and far between. In this tutorial, we will discuss a number of applications that have been successful, contrast them with similar applications using other AI techniques, and look at what makes a successful argumentation application.

Abstract: I will give an overview of recent applications of bipolar argumentation frameworks in explainable AI. The tutorial will be self-contained and does not make any assumptions about previous knowledge, but familiarity with basic abstract argumentation ideas is helpful. I will introduce some classical and quantitative bipolar argumentation ideas and present applications in areas like explainable recommender systems, explainable decision making and explaining machine learning models.

For more info, see Social Program

Abstract: to be added