Synthesizing goal models from declarative data-centric process models

Knowledge-intensive processes progress towards the achievement of operational goals. These processes typically rely on data to enable data-driven decision making, but also require substantial flexibility to deal with the complex and dynamic environments in which they operate. Consequently, declarative data-centric process modeling languages such as the Case Management Model and Notation (CMMN) have been proposed to model knowledge-intensive processes. However, while such process models allow to express goals, they specify dependencies between the goals only implicitly. This makes the goal-oriented behavior of declarative data-centric process models hard to understand, and therefore obfuscates the goal-oriented behavior of knowledge-intensive processes. This paper defines a structural, semi-automated approach to explicate the goal-oriented aspects of declarative data-centric process models. The approach first derives goal relations from a declarative data-centric process model and next synthesizes these goal relations into a goal model using an algorithm. The approach is supported by a tool and has been evaluated in case studies. Using the approach, implicit goal dependencies in declarative data-centric process models are expressed in goal models. This supports the understanding of goal-oriented aspects of declarative data-centric process models.