GAOOLE: a Gaia design of agent-based online collaborative learning environment
Shuangyan Liu, Mike S. Joy, Nathan Griffiths · Open Research Online (The Open University) · 2009
Abstract: An intelligent collaborative learning environment (iCLE) provides an online learning community with an interactive and multi-functional work area with intelligent support for the whole cycle of collaborative education, including organizing teams, advising on group work and communication, tutoring, and testing individual contributions. An agent-based approach lends itself to developing iCLE systems since many of the desired properties and requirements of iCLE systems coincide with those provided by the use of agents, such as autonomy, reactiveness and proactiveness (goal-oriented). Existing agent-based designs for online collaborative learning identify the agent types and the system topology, but lack certain design specifications. In particular, there is a lack of precision with respect to areas such including: the key roles that intelligent agents can play in online collaborative learning management; the computational resources consumed and generated by a role for performing a pedagogical task; the protocols adopted for the interactions between different roles; the agent types with mapped roles and the number of instances of each type in an actual system; and the services that the agents provide. Fully specifying these aspects will enable the system to fully exploit the strengths of agents (including pro-activeness, autonomy and flexibility). In this paper, we propose a new design, GAOOLE (Gaia Design of Agent-based Online Collaborative Learning Environment), which includes a detailed analysis and design specification. It consists of five sub-models: the environment model (describing the computational