LAFP: Logic-based Argumentation Framework with Preferences

Explaining query answers in inconsistent knowledge bases (KBs) has been a prominent problem in AI, especially in Semantic Web. However, in the context of inconsistent KBs, particularly when the date comes with preferences, the problem of explaining query answers remains a challenge, for instance, the explanation can not be exhaustive and user-friendly. We aim to provide a generic and unifying framework in which we use argumentation as an explanation tool. To do so, we first translate inconsistent prioritized KBs to argumentation frameworks (AFs). We then investigate how the AF can provide persuasion dialogues constructed from argumentation-based explanations. As an initial step of our work, the translation of prioritized KBs into AFs is crucial before providing persuasion dialogues. This work has been studied in our submitted paper. Therefore, we place ourselves to explore the problem of translation of inconsistent prioritized KBs into AFs. We propose a novel logic-based argumentation framework with preferences (LAFP). We also provide a theoretical analysis of the LAFP with respect to its syntax and its properties. Finally, we empirically study the applicability of our proposed framework in real-data applications. The experimental results indicate that the proposed one provides more comprehensive explanations for query answers. It is worth mentioning that this paper is a technical report of our pre-release version, and we will report our specific evaluation for the experiment in future versions