The Integrating Modelling Architecture (IMA) is the middleware that enables many of the Ecoinformatics Collaboratory's projects and one of the main areas of development here. It is a framework where all concepts used in modelling of natural systems are explicitly defined by standard ontologies. The IMA derives its power and generality from adopting the semantics of the natural entities represented (e.g. economic value, biomass, or nitrogen flow) as opposed to that of the workflow or model that calculates the desired results. It is not committed to a particular set of modelling concepts except what expressed by its core (upper) ontologies, which are carefully designed for maximum generality, paradigm independence, and uncoordinated extensibility. Many common modelling paradigms and toolkits can be implemented in the IMA by defining new ontologies in the terms of the core ones and - when necessary - providing software extensions in the form of shared libraries to complement the functionalities of an IMA runtime system.

The IMA is a conceptual framework and software design, which is implemented in the Integrating Modelling Toolkit, an open source software package being developed at the Ecoinformatics Collaboratory. Both the design and the software are used in several large projects worldwide.

The IMA design is large and complex, and the project has its own web site at www.integratedmodelling.org. Please refer to this site for in-depth information and documentation. Among the main applications of the IMA, you can check out our Ecosystem Services Database.

For a quick introduction to the IMA concepts, please see the following draft paper:

Villa, F. A semantic framework and software design to enable the transparent integration, reorganization and discovery of natural systems knowledge. Journal of Intelligent Information Systems, in press. Draft: ecoinformatics.uvm.edu/papers/villa_jiis.pdf.