Colloquium 1

The Thrill of Discovery: Information Visualization for High-Dimensional Spaces
Ben Shneiderman, University of Maryland (10/04/07)

As people living in the 'Information Age', we are dealing with more, almost overwhelming, information than ever. Since computers enforce us to store stacks of data not only from the present but from the past as well, it is likely that the amount of data we have will keep increasing as time goes. Therefore, it is important to find some way to manage these data so we can get the best out of them.

When the data is shown in numbers, it is hard to interpret, search necessary items, or find patterns out of it. Since human image storage is fast and vast: in terms of computer science, human visual bandwidth is enormous, visualizing information can help to process the data in many ways.

Dr. Shneiderman described that there are 4 steps to design Graphical User Interface(GUI) for Information Visualization: overview, multiple selection, zoom & filter, and details on demand. Most visualizations are done in 2 dimensions(2D) but there are cases that 3D becomes useful such as world-map, etc. As a demonstration, he used 'Treemap' developed at the University of Maryland.

'Treemap' categorizes data by using three ways: size, color, and group. The biggest advantage of this tool is that you are empowered to discover some information that you didn't even think about from the original text. He also demonstrated 'Time Searcher' for temporal data that enables us to find patterns over time.

Overally, visualizing information makes it easier to find trends or outliers of certain events, and cluster data in more efficient way. The more data we obtain, the more useful this technique will be.