Hypermodelling: an integrated approach to dynamic system modelling

Abstract

Models used in simulation tend to be disconnected from reality and from related models. Ideally, we should like to interact with models as if they were web assets containing extensive hyperlinking and cross-connections. In practice, these inter-connections with respect to the human-model interface are lacking and need to be improved. For example, interacting with a term within a mathematical model is rarely possible. Without such interaction, the full semantics of the term, including its real-world meaning, appearance, and source code implementation, are difficult to identify. The same model may appear in several separate locations: electronic documentation, formula-oriented software, slides, source code. Without the proper connections, model semantics can drift. Also, multiple models are generally required to view different aspects of reality, and to ensure that the right model medium is used for the people that need to understand that model. We introduce the term, hypermodel, to include interaction within models, among models, and between human and model. The discussion of hypermodelling is presented with a special focus on research performed at the University of Florida over the past 20 years.

Panel on grand challenges for modeling and simulation

“…Already new areas are emerging. For example, Hypermodeling, the general theory and practice of linking system models and their components, has been proposed by Fishwick (Fishwick 2012).…”

“…The Semantic Web gives potentially an abstract foundation through the use of ontologies and associated languages to form the structure of an M&S ecosystem on to which actual model elements, data and their relationships can be mapped and studied. Hypermodelling subsumes and extends some approaches to connect and integrate simulation models (Fishwick 2012). Hypermodelling includes issues of model linking, interoperability (to models, system models in other forms, media, heterogeneous model types and custom representations of the same model) and, to some extent, reusability.…”

“…Even though Ezzell et al (2011) attempt to answer some of these questions for the human heart, our community has only begun the scratch the surface for the model interconnection problem in general. Fishwick (2012) presents one framework in "hypermodelling" to capture some of the problems listed in this section. The dominant practice when multiple models are present for model authoring or analysis is to wrap each in a tool, and then integrate the tools.…”

Challenges for web simulation science

“…An M&S ecosystem is "everything"; the emergent complexity of people, organizations and society with regard to M&S artifacts (models, tools, technology, data, etc.). A key component to this is Hypermodeling which subsumes and extends approaches to connect and integrate simulation models (Fishwick 2011). It is strongly associated with the notion of hyperdocument, a media form that interrelates information and media to extend previously static interpretations of documents.…”

“…In Modeling and Simulation (M&S) the Semantic Web may have a significant effect on methods and tools, particularly in model ontologies, discovery, composition, interoperability and reuse. Hypermodeling, the general theory and practice of linking system models and their components, has been proposed by Fishwick (2011). Combining this with Web Science arguably gives us the embryonic field of Web Simulation Science (WSS).…”