The complexity of evaluating e-learning has become an issue that is being addressed by educational developers. A possible solution to part of the problem lies in the use of software agents to extract data from the e-learning software and to organise that data in intelligent ways. A project group (PEDANT) is working on the design and development of pedagogical agents to monitor and evaluate a range of computer-based learning tools including Web-based ones operating as an e-learning environment. This paper reports on the conceptual structure that has evolved to define the development process for the pedagogical agents. This includes the intention of making the final products generalisable across a variety of e-learning platforms and environments.
Proceedings of the Exploring Educational Technologies Conference, Melbourne, Victoria, Australia, 16-17 July 2003