This paper reports on a qualitative study of the use of declarative process notations used in a commercial setting. Specifically, we investigate the actual use of a system implemented in terms of DCR graphs for the Danish "Dreyer Foundation" by our industry partner Exformatics A/S. The study is performed by analysing logs of actual use of the system. By the study we seek to illuminate the currently open -- and heavily debated! -- research question of whether the claimed advantages in conciseness and flexibility of declarative models over more traditional flow-based notations are significant in practice. Studies investigating this question have typically focused on understandability for users, driven by lab experiments and workshops with practitioners unfamiliar with the paradigm. In the present study, we (1) attempt to assess qualitatively whether users employed the flexibility granted them by the declarative model, and (2) use process discovery techniques to examine if a perfect-fitness flow-based model representing the main business constraints is in fact easy to come by. For (1), we find evidence in various forms, most notably an apparent change in best practices by end-users allowed by the model. For (2), we find no such model. We leave as a challenge to the community the construction of a flow-based model adequately representing the business constraints and supporting all observed behaviour by the users, whether by hand or by mining.
Title of host publication
Proceedings of the 2015 Symposium on Computational Intelligence and Data Mining (CIDM)
This page is printed from https://en.itu.dk/research/portalplaceholder?layoutfraction=top&langRef=https://pure.itu.dk/portal/da/clippings/this-powerful-ai-technique-led-to-clashes-at-google-and-fierce-debate-in-tech-heres-why(9b7c1092-626c-4846-942c-85808da17b25).html