Article Preview
TopIntroduction
During the last decades, there is a rapidly growing interest on Business Process Management (BPM), revealed by the plethora of relative emerging technologies. Such technologies include business process modeling languages offering the means to design technically interpretable business process models like BPMN (OMG, 2008), EPCs (Mendling 2008) and S-BPM (Fleischmann and Stary, 2012), tools for business process analysis like Signavio (www.oracle.com/technetwork/middleware/bam/overview/index.html).
The central concept behind all these technologies is the business process model, which is designed, analyzed, implemented, enacted, monitored and evaluated continuously in the course of the business process management lifecycle (Weske, 2007). However, in contrast to the wide variety and rapid evolution of such technologies that support the management of a business process model throughout its lifecycle, less progress has been made on how to design a business process model in terms of a well-defined methodology (Nurcan & Schmidt, 2012) and even less on how to elicit the model from the real world cases (Mauser et al., 2009).
According to Weske (2007), elicitation and modeling together correspond to the design stage of the BPM lifecycle. The reason they are not depicted as separate phases is that they are not conducted in a sequential manner. Rather, they are performed interchangeably until a process model is reached that sufficiently reflects the real-world process. Typically, a process analyst, i.e. a person qualified to drive the design and analysis procedures, with competence on one or more process modeling languages, interviews a representative group of workers, in order to extract information on how the work is done for a specific business process. Subsequently, the process analyst uses this information to build an initial process model. Afterwards, the analyst proceeds with its calibration, which involves iterations of comparisons between the model and the actual process through further interviews with the involved participants, and exploitation of the discrepancies between the two, to improve the model. This procedure is repeated until model accuracy is judged to be acceptable. The construction of the model is typically based on hierarchical top-down modeling like the approach presented in (Silver, 2009). This approach starts with an abstract model following the strategies and policies set by top management. The abstract model is gradually refined into a more fine-grained representation of work based again on information acquired from managers. However, during the decomposition procedure, process participants may also come into play.