Visualization Of Cyclic Multivariate Data

MS Thesis Presentation

Benjamin Lipchak
WPI Computer Science Department
Advisor: Prof. Matthew Ward
Reader: Prof. Michael Gennert

Image Science Research Group
Tuesday, November 11, 1997
10 a.m.
Fuller Labs 141

Cyclic multivariate data increasingly dominates a variety of disciplines, including economics, astronomy/astrophysics, medicine, global change studies, and production management. This data is characterized by some cyclic, or periodic, component in one or more of its dimensions. The ability to visualize the data in such a way that its cyclic nature were enhanced and accessible would be highly desirable, as it would empower the viewer to better understand and interact with the data set. Previous techniques either fail to enhance the cyclic structure of the data or do not scale well for use with multivariate data.

We describe a new interactive visualization technique for cyclic data which accommodates data of many dimensions. This technique, called the Glyph Spiral, employs conventional glyphs, such as star glyphs or bar glyphs, positioned on a spiral such that successive cycles radiate from the center. Intra-cycle as well as inter-cycle relationships are preserved. Furthermore, we describe an interface by which the data set may be navigated.

Maintained by webmaster@wpi.edu
Last modified: Sep 27, 2006, 16:05 EDT