ECAI 2004 Conference Paper

[full paper]

Perception Based Cognitive Diagnosis

Antoaneta Serguieva, Tariq Khan

The development of intelligent educational systems faces the challenging problem of cognitive diagnosis. This necessitates the development of approaches for analyzing user performance and inferring cognitive states. We will focus on the transformation of model-based technical diagnosis into model-based cognitive analysis, which is intrinsically interactive. The purpose of cognitive analysis, in contrast to device diagnosis, leads to development of novel approaches to interpreting student performance. We support the view that Zadeh’s computational theory of perception compliments qualitative methodologies by providing an additional level of information granulation and a computational inference engine. The computational theory of perception is considered complimentary to qualitative methods when processing and reasoning with perception-based information, and this perspective will allow us to reformulate performance analysis. The adopted view throughout this article is that concepts in any domain, as well as their interrelations, are better communicated through a variety of models, each providing partial definition or exemplification from a different perspective. The problem-solving knowledge involving each concept will be described with a set of models arranged in a multi-model space along various modelling dimensions. The representation techniques underlying these models will be described in the paper.

Keywords: computer-aided learning, cognitive diagnosis, perception-based reasoning, model-based reasoning, knowledge representation

Citation: Antoaneta Serguieva, Tariq Khan: Perception Based Cognitive Diagnosis. In R.López de Mántaras and L.Saitta (eds.): ECAI2004, Proceedings of the 16th European Conference on Artificial Intelligence, IOS Press, Amsterdam, 2004, pp.1091-1092.