P. Forbrig and F. Paternò (Eds.): HCSE/TAMODIA 2008, LNCS 5247, pp. 41–57, 2008. © IFIP International Federation for Information Processing 2008 From Task to Agent-Oriented Meta-models, and Back Again Steve Goschnick, Sandrine Balbo, and Liz Sonenberg Interaction Design Group, Department of Information Systems, University of Melbourne, 3010, Australia {stevenbg,sandrine,l.sonenberg}@unimelb.edu.au Abstract. In the research discussed here, in addition to extracting meta-models from numerous existing Agent architectures and frameworks, we looked at sev- eral Task meta-models, with the aim of creating a more comprehensive Agent meta-model with respect to the analysis, design and development of computer games. From the agent-oriented perspective gained by examining the resultant extensive agent meta-model – named ShaMAN – we then revisit the Task Analysis research domain, and consider what benefits Task Analysis and Mod- elling may draw from the Agent-oriented paradigm. Keywords: Agent-oriented, Task Models, Multi-Agent Systems, Meta-model, Agent Meta-models, Task Meta-models, Software Engineering, Computer game development, Agents in computer games. 1 Introduction Agent-oriented (AO) architectures and methodologies are the main interest area of the research outlined here, with a focus on the application domain of computer games. In addition to extracting meta-models from numerous existing Agent architectures and frameworks (not covered in this paper), we looked at several Task meta-models, all with the aim of creating a more comprehensive Agent meta-model with respect to the analysis, design and development of computer games. From the agent-oriented per- spective gained by examining the resultant extensive agent meta-model – named ShaMAN – we revisit the Task Analysis research domain, and consider what benefits Task Analysis and Modelling may draw from the Agent-oriented paradigm. 1.1 Motivation for Task Models in AO Meta-model Research A modern sophisticated computer game can be characterised as a mixed-initiative multi-agent system – meaning that interaction happens between the human us- ers/players, and various game-based synthetic characters, which have a high degree of proactive autonomous behaviour. In addition to being multi-agent in nature, such games also involve multiple users, playing alone or in guilds (teams). AO researchers have predominantly been intent on putting intelligence into artefacts (e.g. trading sys- tems, robots, simulations, etc.), with only a small percentage concerned with mixed- initiative human-agent systems [9,11,12], such as computer games.