Grant: Strategic Projects - Group

In recent years, visualization has been gaining importance as a means of managing the often overwhelming amount of digital information available to users. However, despite the many attempts to identify mappings between user needs and effective visualizations, results are still inconclusive and often conflicting. We argue that a reason for these results is that most of these attempts relied on the assumption that optimal visualizations can be defined based solely on the target data/user task, while we believe that individual user differences also play a role. Given a task, different visualizations may work best for different users, depending on user traits such as cognitive skills, knowledge and preferences. In addition, a user's visualization needs may change dynamically as they learn to use a visualization tool or as they experience different mental or affective states. The goal of the proposed research is two-fold. First, we want to develop new principles to define the role of individual differences in different data visualization tasks and lay the foundations for subsequent tool development. Next, we will use these principles to develop adaptive visualizations, i.e. advanced visualization tools that can model and adapt in real time to those user's traits that impact visualization effectiveness. Although the benefits of user-adaptive interaction have been shown in relation to interaction tasks such as operation of menu-based interfaces, web browsing, and information retrieval, these ideas have rarely been applied to data visualization, mostly because there is limited understanding on which user features the visualization should adapt to and how. After we gain this understanding in the initial phases of our project, we plan to demonstrate the benefits of the resulting tools by implementing and evaluating a capstone system for an application domain of interest to one of our industrial partners.