On the discovery of declarative control flows for artful processes

Artful processes are those processes in which the experience, intuition, and knowledge of the actors are the key factors in determining the decision making. They are typically carried out by the "knowledge workers," such as professors, managers, and researchers. They are often scarcely formalized or completely unknown a priori. Throughout this article, we discuss how we addressed the challenge of discovering declarative control flows in the context of artful processes. To this extent, we devised and implemented a two-phase algorithm, named MINERful. The first phase builds a knowledge base, where statistical information extracted from logs is represented. During the second phase, queries are evaluated on that knowledge base, in order to infer the constraints that constitute the discovered process. After outlining the overall approach and offering insight on the adopted process modeling language, we describe in detail our discovery technique. Thereupon, we analyze its performances, both from a theoretical and an experimental perspective. A user-driven evaluation of the quality of results is also reported on the basis of a real case study. Finally, a study on the fitness of discovered models with respect to synthetic and real logs is presented.