Natural user interfaces in 3D data visualization

Abstract

Data visualization is a concept that has been around for centuries. Due to recent advancement in technology, one can easily see this concept being applied heavily in almost every field of work in the forms of computer applications that allows users to interact and view the data as they best see fit. Fluent interfaces for such applications become increasingly important the larger and more complex data gets, something WIMP GUIs have been seen to struggle with. The purpose of this study is to provide an insight on the possible benefits and limitations using natural user interfaces (Also known as POSTWIMP user interfaces) in 3D data visualization applications may provide. A data visualization application that visualizes data as 3D clusters was developed. The application provides 2 user interfaces, one WIMP, mouse and keyboard focused and the other speech and hand gesture recognition focused (POST-WIMP user interface). 15 participants were asked to perform a set of basic data visualization tasks, once with the WIMP user interface and another time with the POST-WIMP user interface and fill questionnaires regarding their experience using both interfaces. Results produced by the performance of the tasks and the questionnaires provided a better insight into the potential effectiveness and current limitations POST-WIMP user interfaces have when used for data visualization. Results showed that tasks performed with the POST-WIMP user interface took less time to be performed. However, participants made more mistakes when using the POST-WIMP user interface. It was shown that biggest contributor to the high error rate was the high learning curve the POST-WIMP user interface has compared to the WIMP user interface. This was further supported by the fact that when performing the final tasks, the participants’ error rates dropped slightly, regardless of the fact that the final tasks were more complex than the earlier ones.