Oct 2005 - Volume 4 Issue 3
Teaching Artificial Intelligence and Intelligent Agents: Challenges and Perspectives
- Johan Kummeneje and Harko Verhagen - Teaching Agent Programming to a Hybrid Student Population
- John Kerins - Developing Critical Insights into Artificial Intelligence
- Owen Cliffe, Marina De Vos, Julian Padget and Edgar Casasola - Teaching Multi-Agent Systems in the UK and in Latin America
- Jörg Denzinger - Teaching Multi-Agent Systems using the ARES Simulator
- José M. Vidal, Paul Buhler and Hrishikesh Goradia - Tools and Lessons from a Multiagent Systems' Class
- T. L. McCluskey and R. M. Simpson - The Use of an Integrated Tool to Support Teaching and Learning in Artificial Intelligence
Editorial by Maria Fasli
University of Essex
Welcome to this special edition of the ITALICS journal of the HEA-ICS on Teaching Artificial Intelligence and Intelligent Agents: Challenges and Perspectives.
Artificial Intelligence (AI) and Intelligent Agents (IA) have been incorporated into the curriculum of Computer Science degree schemes for a number of years now at both undergraduate and postgraduate levels. One of the challenges is that students attending such courses may have diverse backgrounds, interests and different motives for choosing them and invariably different expectations. Moreover, despite the fact that the underlying research areas have developed over the years, teaching artificial intelligence, agents and multi-agent systems presents a number of problems. Firstly, a great diversity in the topics covered as there is a lack of agreement on the core contents of such courses. Secondly, a heavy influence of one's own research expertise and specialization in deciding the content of such courses. Thirdly, a lack of standard methodologies and tools that can be employed for teaching such topics. This special issue contains papers that discuss problems specific to teaching AI and IA and approaches to learning and teaching these topics.
Kerins' paper (Developing Critical Insights into AI) describes an Artificial Intelligence course that attempts to provide insights into AI through introducing context, theory, and relevant practical tasks that allow students to gain a deeper understanding into some of the scientific and engineering goals of AI. The paper by McCluskey and Simpson (Use of an Integrated Tool to Support Teaching & Learning in AI) reports on the advantages of using a high level integrated tool for supporting teaching and learning Artificial Intelligence topics such as knowledge acquisition and formulation, validation and maintenance of domain models, inductive learning and automated plan generation. They argue that the use of such a tool supports teaching and enhances the students' learning experience.
The papers by Jose M. Vidal et al (Experiences Teaching Multi-agent Systems), Kummeneje and Verhagen (Teaching Agent Programming in Sweden) and Owen Cliffe et al (Teaching Multi-Agent Systems in the UK and in Latin America) describe courses on multi-agent systems being taught in the UK, US, Latin America and Sweden. They complement each other in that they enlighten about practices, experiences and lessons learnt from using different tools and approaches in teaching multi-agent systems topics. These papers provide a useful guide to practitioners looking for tools to complement the theoretical part of multi-agent systems courses and provide practical tasks for the students. Moreover, they provide valuable
insights into the use, appropriateness and effectiveness of the reported tools. The paper by Denzinger and Kidney (Teaching Multi-Agent Systems Using the ARES Simulator) looks into the use of a specific tool in a multi-agent systems course which enables students to gain experience with and tackle problems regarding cooperation and competition in multi-agent systems.
The common theme that emerges from all papers is that students gain a deeper understanding of artificial intelligence and intelligent agents when they engage in challenging tasks that provide them with the opportunity to apply and integrate principles and techniques taught as well as their own ideas. They are then in a position to appreciate the aims and goals of AI and multi-agent systems and the great difficulties and challenges that these present.