The incorporation of information visualization (IV) techniques into a system, in particular a decision support system, is aimed to provide decision-makers with better interpretation and greater insights, and to reduce users’ cognitive load. To facilitate in design and evaluation of IV techniques for decision supports, taxonomy of information visualization is required. Reviewed on the IV taxonomies has found that the current classification of IV is too general or more towards technology oriented, which is acceptable for proficient designer but not for the purpose of evaluation particularly for the business decision support application. Further exploration is required to identify the dimensions that are benefited and crucial in information visualization design and evaluation.
Taxonomy is a model of knowledge to represent part of explanatory theory. It classifies information objects to facilitate discovery of knowledge. The significance of taxonomy has been noted by Simon who argued that “an early step towards understanding any set of phenomena is to develop taxonomy”.
To facilitate in design and evaluation of IV techniques for decision supports, this paper proposes, a novel Expressive IV Taxonomy. It emphasizes on the human aspects, and addresses the question on what dimensions of IV can influence and enhance human cognitive capabilities for better interpretation and analysis of data. More specifically, this taxonomy focuses on dimensions that provide impacts on the expressive quality in visualizing information for decision support. This taxonomy identifies four dimensions namely, representation, organization, content, and interaction. These dimensions are important in designing effective and expressive IV.
An experimental study was conducted to provide an empirical evidence of their importance which is measured by user satisfaction. User satisfaction rating is based on Likert scale range from 1 to 7. A higher mean for user satisfaction indicates that a greater satisfaction is perceived. The results showed that all mean score were higher than 5, with an overall user satisfaction mean score for expression dimensions of 5.25. Therefore, this suggests that these dimensions are important and affects their satisfaction. In addition the results also indicate that there are significant differences across IV techniques on these expressive dimensions. In addition, a significant difference was also found across decision style. This suggests and provides empirical evidence that the proposed dimensions of Expressive IV Taxonomy is important to be considered in the design and evaluation of IV in particular and user interface in general.