Combining quantitative and qualitative reasoning in defeasible argumentation

Labeled Deductive Systems (LDS) were developed as a rigorous but exible method- ology to formalize complex logical systems, such as temporal logics, database query languages and defeasible reasoning systems. LDSAR is a LDS-based framework for defeasible argumentation which subsumes di erent existing argumentation frameworks, providing a testbed for the study of dif- ferent relevant features (such as logical properties and ontological aspects, among others). This paper presents LDS AR, an extension of LDSAR that incorporates the ability to combine quantitative and qualitative features within a uni ed argumentative setting. Our approach involves the assignment of certainty factors to formulas in the knowl- edge base. These values are propagated when performing argumentative inference, o ering an alternative source of information for evaluating the strength of arguments in the dialectical analysis. We will also discuss some emerging logical properties of the resulting framework. Eje: Lógica e Inteligencia artificial Red de Universidades con Carreras en Informática (RedUNCI)