Evaluation of information visualization tools using the NFR approach

Abstract

Determination of quality of information visualization (infovis) tools will assist practitioners in industry and academia to choose the appropriate tools for their needs. Infovis tools are widely used to visualize data-intensive applications so that quick decisions may be taken. In this paper we use a novel technique, the NFR Approach, to determine quality of infovis tools. The NFR Approach considers visualizability as an important non-functional requirement (NFR) for infovis tools and evaluates visualizability to determine the quality of these tools. The chief artifact that the NFR Approach employs for evaluation is the softgoal interdependency graph (SIG) wherein factors affecting visualizabil-ity are related to tool features by means of contributions that come in four flavors. Well-defined propagation rules help determine a qualitative score for visualizability based on the contributions. The chief advantages of the NFR Approach include rationale for the scores, flexibility to accommodate different definitions for visualizability, and historical record-keeping. The NFR Approach is validated by applying the Approach to three infovis tools - Prefuse, Tom Sawyer, and GnuPlot. © 2008 Springer-Verlag Berlin Heidelberg.