Page 353 -

13.2 Process Mining: TomTom for Business Processes? 335

maximized. For every possible next action, the value of the performance indicator is
predicted. This information is used to rank the possible actions and thus recommend
the next step to be taken (cf. Fig. 9.12).
Recommendations based on process mining allow for systems that are flexible
but also supporting operational decision making. Today’s information systems typ-
ically do not provide a good balance between flexibility and support. The system is
either restricting people in their actions or not providing any guidance. BPM systems
offering more flexibility (e.g., case handling systems like BPM|one or declarative
workflow systems like Declare), can be extended with a recommendation service
based on process mining techniques [110].
The TomTom metaphor illustrates that many information systems lack function-
ality present in today’s navigation devices [94]. However, high-quality process mod-
els tightly coupled to event logs enable TomTom-like functionalities such as predict-
ing the “arrival time” of a process instance, recommending the next activity to be
executed, and visualizing “traffic jams” in business processes.