A constraint programming model for making recommendations in personal process management: A design science research approach

Oruç S., Eren P. E. , Koçyiğit A.

DECISION SUPPORT SYSTEMS, vol.152, 2022 (Journal Indexed in SCI) identifier identifier

  • Publication Type: Article / Article
  • Volume: 152
  • Publication Date: 2022
  • Doi Number: 10.1016/j.dss.2021.113665
  • Keywords: Personal process management, Design science research, Constraint programming, Business process management, Intelligent personal assistants, WORKFLOWS


Decision-making in everyday life has an essential role in effectively completing personal tasks and processes. The complexity of these processes and the resulting cognitive load of managing them may vary significantly. To decrease the cognitive load created by such decision-making efforts and to obtain better outcomes, recommendation systems carry significant potential. In order to investigate the benefits provided by decision support systems (DSS) in personal process management (PPM), we first build a constraint programming (CP) model and a prototype context-aware-mobile application employing this CP model. Then, we evaluate the application and the model via two exemplary real-world scenarios. The scenarios form the core of the experiments conducted with 50 participants. We compare the participants' planning performances with and without the PPM system with quantitative metrics such as planning times and scenario objective values. In addition, System Usability Scale (SUS) questionnaires and open-ended questions provide qualitative evaluation results. Throughout the study, we apply the Design Science Research methodology to rigorously conduct research activities by proof of concept, proof of use, and proof of value. The empirical results clearly show that our proposed model for PPM is effective, and the developed prototype solution generates positive participant comments as well as a high SUS score. Overall, the prototype PPM system with CP implementation leads to better planning in less time in the planning phase, and it lets the user do fast replanning in the execution phase, which is invaluable in dynamically changing situations such as daily activities.