A framework for estimation of complexity in agent oriented methodologies

Agent-oriented software methodologies have found their way to become one of the most important issues of software development. The diversity of existing methodologies and the variety of newly arriving ones have arisen the question of which methodology is best in the context of a special problem, so the need for evaluating and comparing methodologies is becoming more important everyday. On the other hand evaluating methodologies can reveal their strengths and shortcomings which lead to an improvement in development of the next generation of agent-oriented methodologies. In this paper, we are going to introduce an approach for evaluating methodologies via the evaluation of their lifecycle process, models and artifacts produced. We show the usage of our method by evaluating three methodologies (MaSE, Prometheus and MESSAGE) and a typical case study. Results provide us with much useful information on strengths and weaknesses of each methodology.