Patterns Of Process Modeling

Wil M. P. Van Der Aalst,Arthur H. M. Ter Hofstede,Marlon Dumas
DOI: https://doi.org/10.1002/0471741442.ch8
2005-01-01
Abstract:9.1 INTRODUCTION The previous chapters have presented different languages and approaches to process modeling. In this chapter, we review some issues in process modeling from a more language-independent perspective. To this end, we rely on the concept of pattern: an " abstraction from a concrete form which keeps recurring in specific non-arbitrary contexts " [18]. The use of patterns is a proven practice in the context of object-oriented design, as evidenced by the impact made by the design patterns of Gamma et al. [10]. Process Aware Information Systems (PAISs) address a number of perspectives. Jablonski and Bussler [11] identify several such perspectives in the context of work-flow management. These include the process perspective (describing the control-flow), organization perspective (structuring of resources), data/information perspective (to structure data elements), operation perspective (to describe the atomic process elements) and integration perspective (to " glue " things together). 1 In a typical work-flow management system, the process perspective is described in terms of some graphical model, e.g. a variant of Petri nets (see Chapter 7), the organization perspective is described by specifying and populating roles and organizational units, the data/information perspective is described by associating data elements to work-flow instances (these may be typed and have a scope), the operation perspective is described by some scripting language used to launch external applications, and the integration perspective is described by some hierarchy of processes and activities. In principle, it is possible to define design patterns for each of these perspectives. However, the focus of this chapter is on patterns restricted to the process (i.e. control-flow) perspective. This perspective is the best understood as well as the dominant perspective of workflow. Patterns for the data perspective have been reported in [20], 1 Note that in [11] different terms are used.
What problem does this paper attempt to address?