Unpacking the argument : A claim-centric view on abstract argumentation

Abstract

The representation of conflicting scenarios in terms of abstract arguments and attacks has been considerably promoted by the work of Dung; his abstract argumentation frameworks (AFs) are a key formalism in AI research nowadays. Claims are an inherent part of each argument; they substantially determine the structure of the abstract representation. While fundamental properties of argument acceptabiltiy are well understood, only little is known about structural and computational aspects of claim acceptability. Moreover, the strict focus on arguments in the abstract representation restricts the modeling capacities of AFs to problems that do not involve claims in the evaluation. In this thesis, we address these issues by unpacking the abstract arguments: we utilize extensions of the abstract model that keep track of parts of the inner structure of arguments. A prominent role in this thesis play claim-augmented argumentation frameworks (CAFs) which extend AFs by assigning a claim to each argument. As we will see, this minimal modification will be sufficient to analyze claim acceptance at a very general level. Inspired by certain shortcomings of traditional claim assessment methods, we propose a hybrid approach to evaluate claim acceptance by shifting certain evaluation steps on the level of claims. We analyze principles and expressiveness of claim semantics and study the computational complexity of conclusion-focused reasoning. We furthermore investigate claim acceptance in a dynamic setting; for this, we extend the model and consider not only the claims, but also the vulnerabilities of the arguments. In summary, this thesis provides a thorough analysis of fundamental properties of abstract argumentation from the perspective of the claims of the arguments.