operating a mouse.
We assumed that the average task execution times for the 2D-GUI and 3DGUI takes were equal and did a non-ordered sign test with the significance level of 0.05. Table 1 shows the results. In Table 1, the circle shows the GUI system which is significantly shorter in average task execution time; the X mark shows the GUI which is significantly longer; and the dashed mark shows the others. Graph 1 shows the average task execution times.
As for task execution time, the 3D-GUI system was significantly shorter when the number of nodes exceeded 100. We got a similar result from questionnaires. This is because when the number of nodes exceeded 100, the 2D-GUI could not display all of the data at one time. In other words, it is generally appropriate to use a 3D-GUI system when all the data can not be displayed at one time with a 2D-GUI and on a scale in which we can recognize the parameter values. Conversely, it is not necessary or appropriate to use a 3D- GUI system when all values can be displayed at one time using a 2D-GUI.
In this paper, we have proposed a effective method for constructing 3D-GUIs semi-automatically. We have also conducted experiments in which human subjects rated 2D-GUIs against 3D-GUIs both of which were generated by our method. These experiments show the effectiveness of' 3D-GUIs. However, in our current version of the system, the automatically created 3D-GUIs are still simple and domain-free parts, and the users are required to customize the generated parts to some extent. We will need to expand our method further in order to be implemented to generate domain-specific and more sophisticated parts automatically.
Misayo Kitamura, and Akira Sugimoto. ( 1995). GhostHouse: A Class Library for Generating Customizable Graphical User Interface. Trans. on Information Processing Society Japan, Vol. 36, No.4, pp.944-958.
Paul S. Strauss, and Rikk Carey. ( 1992). An object-oriented 3D graphics toolkit. Proceeding of SIGGRAPH̕92, pp.341-349, ACM Press.
Josie Wemecke. ( 1994). The Inventor Mentor. Addison Wesley.
K. M. Fairchild, S. E. Poltrock, and G. W. Furnas. ( 1988). SemNet: Threedimensional graphic representation of large knowledge bases. Cognitive Science and its Applications for Human-Computer Interaction, pp.201-233, Lawrence Erlbaum Associates.