Assumption-based Argumentation (ABA) frameworks have been advocated as unifying frameworks for various forms of non-monotonic reasoning, including logic programming. In recent work we have pre- sented an approach for learning ABA frameworks from background knowl- edge and positive and negative examples. Our approach is based on the use of transformation rules, including some adapted from logic program transformation rules (notably folding) as well as new, specific ones, such as rote learning and assumption introduction. In this paper we address the problem of automating the learning of ABA frameworks in the case where acceptance is defined in terms of brave reasoning under stable ex- tensions. We present an algorithm for applying the transformation rules and its implementation that makes use of Answer Set Programming
Learning Brave Assumption-Based Argumentation Frameworks via ASP
Emanuele De Angelis;Maurizio Proietti;
2023
Abstract
Assumption-based Argumentation (ABA) frameworks have been advocated as unifying frameworks for various forms of non-monotonic reasoning, including logic programming. In recent work we have pre- sented an approach for learning ABA frameworks from background knowl- edge and positive and negative examples. Our approach is based on the use of transformation rules, including some adapted from logic program transformation rules (notably folding) as well as new, specific ones, such as rote learning and assumption introduction. In this paper we address the problem of automating the learning of ABA frameworks in the case where acceptance is defined in terms of brave reasoning under stable ex- tensions. We present an algorithm for applying the transformation rules and its implementation that makes use of Answer Set ProgrammingI documenti in IRIS sono protetti da copyright e tutti i diritti sono riservati, salvo diversa indicazione.