Architectural modeling of metamemory judgment in case-based reasoning systems

This paper presents an approach to model metamemory judgments in Case Based Reasoning (CBR) systems. Initially are described the theoretical references about metamemory judgment and CBR. Then it is presented the modeling of metacognitive skills, related to metamemory and a kind of software architecture for general models of CBR called MJ-CBR. This architecture can provide to CBR system with the metacognitive ability to make judgments about the own learning process of a new case. To validate the architecture was developed an Intelligent Tutoring System (ITS) founded in CBR and discovery learning, denominated Learning Mandarin. Finally are presented the empirical results of MJ-CBR performance.