On the Incremental Computation of Argumentation Frameworks

We tackle the problem of efficiently recomputing extensions of dynamic Argumentation Frameworks (AFs) under well-known semantics, namely frameworks able to model human ability to argue. We introduce an incremental approach that, given an initial argumentation framework, an initial extension for it, and an update, computes an extension of the updated framework. The proposed approach is then extended to different kinds of argumentation frameworks (i.e., abstract, bipolar, extended, structured), and is able to incorporate existing AF-solvers in order to compute an extension of the updated framework.