Abstract

Abstract. Goal- and agent-oriented models have emerged as a way to capture stakeholder and organizational goals in a complex enterprise. The complexity of such models leads to a need for systematic procedures to enable users to evaluate and compare the alternative actions and solutions expressed in models. Many existing approaches focus on automated procedures, limiting the ability of the user to intervene. Here, we introduce a qualitative, interactive evaluation procedure for goal- and agent-oriented models, allowing the modeler to supplement the evaluation with domain knowledge not captured in the model. We provide a sample methodology to guide model creation and domain exploration which includes the evaluation of alternatives. We illustrate the procedure and methodology with the i * Framework. Case study experience shows that the procedure facilitates analysis, prompts iteration over model development, promotes elicitation, and increases domain understanding. We describe the results of an exploratory experiment designed to test these findings. 1

Citations