Diagrams 2000 - Tutorial Programme

Formal Approaches to Visual Language Specification and Understanding Kim Marriott School of Computer Science and Software Engineering Monash University, Australia Email: marriott@csse.monash.edu.au Two of the most fundamental questions in visual language research are how to specify a visual language and how to recognize and understand diagrams in a particular visual language. In this tutorial we survey the many formalisms which have been suggested over the last three decades for visual language specification, discuss computational approaches to diagram understanding based on these formalisms and indicate possible applications. We shall also review recent directions in visual language theory, notably efforts to develop an analogue of the Chomsky hierarchy for visual languages, the specification of diagrammatic reasoning, and cognitive models of visual language understanding.

Cognitive History of Science: the Roles of Diagrammatic Representations in Discovery and Modeling Discovery David Gooding Science Studies Centre, Department of Psychology University of Bath, UK Email: hssdcg@bath.ac.uk This session looks at some uses of diagrams in scientific discovery, particularly their role as intermediate representations which mediate between phenomena, descriptions which can be communicated and descriptions which are general. A range of examples will illustrate a variety of uses, including: the abstractive, generative role; diagrams as encoded knowledge; reasoning with diagrammatic representations in discovery; and communication (exposition and argumentation). The tutorial will encourage consideration of two issues: (a) whether, from a cognitive standpoint, diagrams are essential to reasoning about natural phenomena and processes, and (b) the relationship of diagrammatic reasoning to other types of visualisation and visual thinking in the sciences, including cognitive and computational modeling of discovery.

Cognitive (production system) modelling of how an expert uses a Cartesian graph Hermi Schijf Dept. of Educational Sciences Utrecht University, The Netherlands Email: h.schijf@fss.uu.nl This tutorial covers, in brief, the road from observing behavior to the implementation of the observed behavior in a mixed rule-based and parallel network computer model. The emphasis will be on production system, or rule-based modelling. Rule-based modeling uses independently firing if-then rules to capture behavior. Why do this type of modeling, what types of data do you need, what are some advantages and limitations of the method? Simple examples of rule-based modeling will be given; these will be extended to a brief explanation of the mixed model. This model gives a theoretical explanation of the behavior of an expert using visual reasoning based on a Cartesian graph combined with verbal reasoning to teach Economics principles to students.

The Coordination of external representations and internal mental representations in display-based cognitive tasks Jiajie Zhang Department of Health Informatics University of Texas at Houston Email: Jiajie.Zhang@uth.tmc.edu Many cognitive tasks, whether in everyday cognition, scientific practice, or professional life, are distributed cognitive tasks--tasks that require integrative, interactive, and dynamical processing of information retrieved from internal representations and that perceived from external representations through the interplay between perception and cognition. The representational effect is the ubiquitous phenomenon that different representations of a common structure can generate dramatically different representational efficiencies, task complexities, and behavioral outcomes. A framework of distributed representations is proposed to account for the representational effect in distributed cognitive tasks. This framework considers internal and external representations as two indispensable components of a single system and suggests that the relative distribution of information across internal and external representations is the major factor of the representational effect in distributed cognitive tasks. A representational determinism is also proposed--the form of a representation determines what information can be perceived, what processes can be activated, and what structures can be learned and discovered from the specific representation. Applications of the framework of distributed representations will be described for three domains: problem solving, relational information displays, and numberation systems.