Article Preview
TopIntroduction
In light of a rapidly increasing need for high-quality “lightweight formal models”1 (e.g., process models, formal ontologies, business rules, and so on) to fulfill the technology-based promises of information systems and AI―including the Semantic Web (Berners-Lee, Hendler, & Lassila, 2001)―the lack of operational methods for formal modeling and, as an embodiment of such methods, tooling to support them, is becoming a problem. The increasing need for truly collaborative modeling can be added to this (de Moor, 1999).
Current tools for modeling are mostly editor-like, technical environments that at best offer some automated model checking, versioning, and file management. Contrary to what many seem to believe or claim, even advanced graphical editors for, for example, UML and BPMN schemas (Booch, Rumbaugh, & Jacobson, 1998; OMG, 2006) still require technically skilled and above all experienced people to wield them successfully. Beyond editing, very little real support for the interactive process of collaborative modeling is offered, in particular if such a process is to be carried out by relatively inexpert participants. As argued at length in Hoppenbrouwers (2008), this is not an acceptable situation in the long run, mostly because experts (modelers, facilitators) in formal modeling are relatively few and expensive. Lightweight, collaborative formal modeling will have to be brought to the masses, somehow. Creating interactive, low-threshold digital environments seems to be a highly promising way of enabling this. The image of “modeling wizards” presents itself. However, such tools simply do not exist at the moment. Creating them involves both the setting and the answering of a score of research questions and requires a long-term effort.
This article aims to be part of setting the scene for such an effort, and to make some aspects more concrete, without any pretense to provide definitive answers.