This paper presents a new interactive scatter plot visualization for multi-dimensional data analysis. We apply Rough Set Theory (RST) to reduce the visual complexity through dimensionality reduction. We use an innovative point-to-region mouse click concept to enable direct interactions with scatter points that are theoretically impossible. To show the decision trend we use a virtual Z dimension to display a set of linear flows showing approximation of the decision trend. We conducted case studies to demonstrate the effectiveness and usefulness of our new technique for analyzing the property of three popular data sets including wine quality, wages and cars. The paper also includes a pilot usability study to evaluate parallel coordinate visualization with scatter plot matrices visualization with RST results.
History
Publication title
Journal of Visual Languages and Computing
Volume
33
Pagination
13-23
ISSN
1045-926X
Department/School
School of Information and Communication Technology
Publisher
Academic Press
Place of publication
United Kingdom
Rights statement
Copyright 2015 Elsevier Ltd.
Repository Status
Restricted
Socio-economic Objectives
Expanding knowledge in the information and computing sciences