Finding Solutions in Goal Models: An Interactive Backward Reasoning Approach

Modeling in the early stage of system analysis is critical for understanding stakeholders, their needs, problems, and different viewpoints. We advocate methods for early domain exploration which provoke iteration over captured knowledge, helping to guide elicitation, and facilitating early scoping and decision making. Specifically, we provide a framework to support interactive, iterative analysis over goal- and agent-oriented (agent-goal) models. Previous work has introduced an interactive evaluation procedure propagating forward from alternatives allowing users to ask "What if?" questions. In this work we introduce a backwards, iterative, interactive evaluation procedure propagating backward from high-level target goals, allowing users to ask "Is this possible?" questions. The approach is novel in that it axiomatizes propagation in the i* framework, including the role of human intervention to potentially resolve conflicting contributions or promote multiple sources of weak evidence.