6/13/2001 1 Using Perceptual Syntax to Enhance Semantic Content in Diagrams Pourang Irani Visual Interfaces Laboratory University of New Brunswick Box 4400 Fredericton, NB Canada E3B 2A3 (506) 453-4566 pirani@nbnet.nb.ca Colin Ware Data Visualization Research Lab Center for Coastal and Ocean Mapping University of New Hampshire Durham, NH 03824 (603) 862-1138 colinw@cisunix.unh.edu Maureen Tingley Department of Mathematics University of New Brunswick Box 4400 Fredericton, NB Canada E3B 2A3 (506) 458-7343 maureen@math.unb.ca Diagrams are essential in documenting large information systems. They capture, communicate and leverage knowledge indispensable for solving problems and are conceived to act as "cognitive externalizations". 1 A diagram provides a mapping from the problem domain to the visual representation by supporting cognitive processes that involve perceptual pattern finding and cognitive symbolic operations. 2 However not all mappings are equivalent, and to be effective a diagram's representation needs to be embedded with characteristics such that meaningful patterns can be easily perceived. Consequently a diagram's effectiveness depends, to some extent, on how well it is constructed as an input to our visual system. 3,4 In our research, we focus on a class of diagrams commonly referred to as graphs or node- link diagrams. Nodes representing entities, objects or processes, and links or edges representing relationships between the nodes characterize them. Their most common form is that of outline circles or boxes denoting nodes, and lines of different types representing links between the nodes. Entity-relationship diagrams, software structure diagrams and data flow models are examples of node-link diagrams used to model the structure of processes, software, or data. Currently the most widely used graphical language for modeling complex systems is the - Unified Modeling Language (UML). UML contains a suite of diagramming techniques that allow one to model various aspects of a software system, 5 a real-time application, 6 or an enterprise structure. 7 Its versatility in several application areas results from the rich semantics it seeks to model. For example class diagrams in UML model software structures and include methods for depicting inheritance and composition. When these semantics are used in the realm of enterprise modeling for example, UML can capture relationships between organizations or relationships between the corporation and its employees. However, although considerable attention has been given to making these UML notations general and complete, the actual choice of graphical notations appears to be somewhat arbitrary; only an expert in the field can easily read them. In this paper we first discuss aspects of structured object recognition theory and show how this can be used to make 3D diagrams that are more easily analyzed and remembered. We present the results of a new set of studies suggesting how careful mapping of problem semantics to 3D diagram structures can make the meaning of the diagrams easier to “read” with minimal training.