Presenter: Brent M. Dennis
Advisor(s): Christopher G. Healey
Author(s): Brent M. Dennis
Graduate Program: Computer Science
Title: Assisted Navigation in Large Information Spaces
Abstract: Modern technology has enabled researchers to collect vast amounts of
information in an expanding scope of research fields. At the same
time, these new datasets are becoming more complex as evidenced with
their increasing sizes and dimensionalities. Managing and
understanding these datasets has become a challenging problem.
Visualization attempts to address these concerns by creating
meaningful graphical representations of data that can rapidly and
accurately convey important information about the data to a
researcher. In order to achieve this, visualizations need to
accurately represent the global structure of a dataset while
simultaneously presenting local detail about individual data elements.
However, many existing visualization techniques can become overwhelmed
with today's datasets. As a result, information is often forced
off-screen due to a lack of visual resources for representing the
entire dataset. We have developed a navigation assistant to
support users with locating possibly important off-screen information.
Our navigation assistant first builds a user preference model to identify a set of elements of interest to the user. These elements are then clustered into spatially coherent areas of interest. The assistant then imposes an underlying graph framework within each area of interest and a global graph framework between the set of areas of interest. This framework can then be used to provide navigational cues to the user for locating off-screen elements of interest. Moreover, this graph framework can be used to construct informative animated tours of individual areas of interest or the entire dataset itself. Using these techniques, the navigation assistant provides an effective tool for exploring the visualization of a complex information space.