From the Behavior Model of an Animated Visual Language to its Editing Environment Based on Graph Transformation

Abstract

Animated visual models are a reasonable means for illustrating system behavior. However, implementing animated visual languages and their editing environments is difficult. Therefore, guidelines, specification methods, and tool support are necessary. A flexible approach for specifying model states and behavior is to use graphs and graph transformations. Thereby, a graph can also represent dynamic aspects of a model, like animations, and graph transformations are triggered over time to control the behavior, like starting, modifying, and stopping animations or adding and removing elements. These concepts had already been added to Dia-Meta, a framework for generating editing environments, but they provide only low-level support for specifying and implementing animated visual languages; specifying complex dynamic languages was still a challenging task. This paper proposes the Animation Modeling Language (AML), which allows to model behavior and animations on a higher level of abstraction. AML models are then translated into low-level specifications based on graph transformations. The approach is demonstrated using a traffic simulation.