Article,
A Principled Taxonomy of Software Visualization
Journal of Visual Languages & Computing, 4 (3): 211--266 (September 1993)
DOI: 10.1006/jvlc.1993.1015
Abstract
In the early 1980s researchers began building systems to visualize computer programs and algorithms using newly emerging graphical workstation technology. After more than a decade of advances in interface technology, a large variety of systems has been built and many different aspects of the visualization process have been investigated. As in any new branch of a science, a taxonomy is required so that researchers can use a common language to discuss the merits of existing systems, classify new ones (to see if they really are new) and identify gaps which suggest promising areas for further development. Several authors have suggested taxonomies for these visualization systems, but they have been ad hoc and have relied on only a handful of characteristics to describe a large and diverse area of work. Another major drawback of these taxonomies is their inability to accommodate expansion: there is no clear way to add new categories when the need arises. In this paper we present a detailed taxonomy of systems for the visualization of computer software. This taxonomy was derived from an established black-box model of software and is composed of a hierarchy with six broad categories at the top and over 30 leaf-level nodes at four hierarchical levels. We describe 12 important systems in detail and apply the taxonomy to them in order to illustrate its features. After discussing each system in this context, we analyse its coverage of the categories and present a research agenda for future work in the area.
