Audio-visual data mapping for GIS-based data: an experimental evaluation

In this work, we present our experience of utilizing audio-visual data mappings for GIS-based information visualization. The application we choose is a GIS-based system for visualizing crime in a city. In this application, we enhance the pseudo-colored visual presentation of crime information by mapping data to several sound parameters — volume, balance, bass and treble. Our motivation for choosing sound in addition to vision is guided by our belief that data quantities mapped to various colors in a coloring scheme do not always clearly describe the information being presented for many different tasks that visualization is expected to support In many cases additional data characteristics can be conveyed to the user through sound to enhance the performance of the user on those tasks. We have conducted experiments with human users to compare the performance of users on visual data mapping alone vs. visual and sound data mappings together on several tasks including estimates of raw data values, local averaging, and global comparison. In most cases, we found that the use of bi-modal visual and sound data mappings together provided more accurate understanding of data displays