Design and evaluation case study: evaluating the kinect device in the task of natural interaction in a visualization system

Abstract

We verify the hypothesis that Microsoft’s Kinect device is tailored for defining more efficient interaction compared to the commodity mouse device in the context of information visualization. For this goal, we used Kinect during interaction design and evaluation considering an application on information visualization (over agrometeorological, cars, and flowers datasets). The devices were tested over a visualization technique based on clouds of points (multidimensional projection) that can be manipulated by rotation, scaling, and translation. The design was carried according to technique Participatory Design (ISO 13407) and the evaluation answered to a vast set of Usability Tests. In the tests, the users reported high satisfaction scores (easiness and preference) but, also, they signed out with low efficiency scores (time and precision). In the specific context of a multidimensional-projection visualization, our conclusion is that, in respect to user acceptance, Kinect is a device adequate for natural interaction; but, for desktop-based production, it still cannot compete with the traditional long-term mouse design.