Towards a Framework for Collaborative Modeling and Simulation

We identify four methodologies that need to be present for modeling and simulation in a collaborative analysis or design project. The first of the four, methodology M1, refers to how participants perceive their problem situation and create static, or descriptive models. This paper describes the requirements for a successful methodology M1. It then describes Conversational Modeling (CM), a software-supported technique for collaborative modeling, and makes an argument for why CM fulfills those requirements. Finally it describes a case study in which CM enabled participants to construct static knowledge models in collaborative sessions. Four methodologies in collaborative modeling The analysis and design of a work system by a collaborative design team using modeling tools can be described as a holon in terms of Soft Systems Methodology [1]. Characterizing modeling as a holon is a way of looking at the whole modeling effort as a systemic process. Figure 1 shows that this modeling process is not as simple as having one methodology for modeling, but in fact there is an interplay of four 1 “ The abstract idea of a whole having emergent properties.” methodologies (M1 through M4) that together define a modeling methodology as a whole. Each methodology defines a purposeful activity system, which itself can be seen as a holon. Methodology M1 — A methodology for perceiving a system and constructing static models A design team needs to have a principled and structured methodology for the process of understanding the problem situation or system as it exists. We describe this methodology, M1, more completely below. Methodology M2 — A methodology for participatory design of simulation models Methodology M2 describes the process by which formal system modelers co-develop a simulation model of a system with the design team members. Methodology M3 — A methodology for implementing, debugging, and validating a simulation model Methodology M3 is very tool-specific, and is focused on the creation and validation of dynamic simulation models. As part of this methodology there are two issues at hand: