CoDe: a Graphic Language for Complex System Visualization

The concept of macroscope has been exploited in several contexts [dR, f] to outline some methodologies aimed to describe the information of complex systems. The basic idea is that complex systems should be looked at as an integrated whole, rather than be described as an aggregate of standalone components. Following this approach, we introduce a Graphic Language named CoDe (Complexity Design) to describe visual representations which bring information according to the basic idea of a macroscope. The syntactic and semantic specification of CoDe is a framework for a methodology aimed to support visual representation of complex information. An extensible set of graphical patterns are the lexical elements of the language, which provides two families of operators: a first set of operators is supplied to compose basic patterns to form complex ones; a second set of operators allow the designer to specify the semantics of graphs by creating a logical link with the data they represent. Thus, the CoDe Graphic Language is suited to implement the features of such a methodology. CoDe is proposed as an interactive tool that support the designer of a visualization project in describing the complex system to be represented, by means of component abstraction and relation discovery, which is the basis for most decision processes. An application to graphic visualization realized by the research group DensityDesign at Politecnico di Milano - INDACO Department [dd] is provided. Some future applications to Data Mining and Datawarehouse are also outlined.