A two-layered multi-agent framework for providing different forms of assessment in a collaborative learning environment

This paper proposes a framework for providing different forms of assessment. This is a generic architecture that possesses three distinct advantages. It can be seen as a general standard framework of assessment. Educational researches can benefit from having such an integrated basis for comparative analysis of different forms of assessment Also it can be used as the basic framework for building an adaptive assessment system. The proposed framework is implemented in an actual assessment system for primary arithmetic operations with Zeus agent building toolkit.