| Using Spotfire: Search Criteria on Efficiency of Dynamic Query Results |
Ivan Chou - ichou@wam.umd.edu
Li-Chung Liu - lcl@glue.umd.edu
Spotfire is a "visual data mining" program which allows users to browse and analyze data in a direct manipulation interface. Users have at their disposal a 2-D starfield display, which visually represents two characteristics (e.g. length, subject) of the data. The characteristics not shown are manipulated through query devices. For datasets with many characteristics, we wanted to see which starfield display characteristics yield the fastest query time. Different color codings on the starfield display were examined, to see the effects on performance, if any. The results were somewhat inconclusive. Subjects appear to have demonstrated a learning effect as they participated in the experiment, skewing results. However, subjective satisfaction and the effect of color codings indicate some useful findings.
Last updated: 5/3/98