Analysis of models and metacognitive architectures in intelligent systems

Abstract

Recently Intelligent Systems (IS) have highly increased the autonomy of their decisions, this has been achieved by improving metacognitive skills. The term metacognition in Artifi cial Intelligence (AI) refers to the capability of IS to monitor and control their own learning processes. This paper describes different models used to address the implementation of metacognition in IS. Then, we present a comparative analysis among the different models of metacognition. As well as, a discussion about the following categories of analysis: types of metacognition architectural support of metacognition components, architectural cores and computational implementations.