The Interaction and Visualization Laboratory (IVL) operates as part of the CECS. The lab has projects in Foundations, Systems and Applications of visualization.
FoundationsFoundations projects seek to extend the scope of knowledge in the fields of graphics and visualization. These projects often have no driving application, but are techniques which may enjoy widespread use in the future.
Data Structure Layouts - Presenting clear pictures of data is one of the core aims of information visualization. Given data organizations such as lists, hierarchies or relational graphs, is it possible to develop broadly applicable display techniques? This project investigates core layout algorithms and their application in various domains. We have developed several new layouts of hierarchical information and are currently working to improve spring embedder graph layout algorithms.
Non-Linear Display Mapping - Presenting large quantities of
data in an understandable format is another core aim of
visualization. Often the best graphical organization is larger than
the available display area, necessitating the use of scrollable
windows or other page at a time techniques. The application of
non-linear scaling can integrate the detail provided by scrolling with
the overview provided by scaling. We have developed several
techniques for non-linear mapping and have applied these techniques to
lists, hierarchies and graphs. The image on the right shows the
combination of these two projects-the image represents a semantic web
of concepts and the user is currently focused on the area immediately
around the "service" node.
Redesigning Familiar Layouts - There are a number of
well-known data presentation formats such as the histogram or line
graph which are a familiar to almost all computer users. These types
of data drawings evolved as a static presentation primarily targeted
toward the printed page. This project seeks to improve on these
established layouts by enhancing them with more data elements
(e.g. increasing the data to ink ratio) and by providing interactive
data exploration within data graphics. The image on the left is the
familiar bar chart enhanced to show the relative values of the
elements in each histogram bin.
The systems area develops tools and techniques to allow better use of computational resources. These types of projects are aimed at assisting high-end users, programmers and system administrators through applications of graphics, visualization and interaction.
Host Monitoring - StatMon is a system
for monitoring large
numbers of computers in a networked environment. The
visualization component of this project is attempting to design
and develop t a glance displays of system status, showing the user
an overview of the systems and recent history for one system.
The board can be minimized to occupy just a corner of the display
area, allowing the user to focus on other tasks while maintaining
awareness of system status. This project is currently under
further development to allow interactive host control and viewing
of machine history.
Algorithm Visualization - We have developed instrumentation
and visualization tools for ParInt, a parallel
integration engine developed at Western Michigan University. This
post-mortem visualization represents region subdivision information
(the key algorithmic component of the computation) in a variety of
displays including abstract representations such as subdivision trees
as well as direct mappings of values such as region division
plots. These tools use SDDF as the data storage mechanism and can be
utilized for other systems as long as the programs are instrumented to
record appropriate run tracing events.
CMM Evaluation - The accurate reproduction of color across various digital devices (cameras, scanners, displays, projectors, printers, etc is a key component of imaging. Color Management Modules utilize color profiles to translate between device color spaces created by either device manufacturers or specialized measurement devices. In close collaboration with the Department of Paper Science, this project seeks to evaluate the efficiency and accuracy of these modules through the design and development of test metrics and evaluation software.
ApplicationsThe IVL develops applications and tools to assist end users with a variety of tasks. These types of projects are sometimes motivated by foundation research, although just as often application development drives the techniques. Direct application development is a valued activity of the lab, assisting both students (as a learning experience) and faculty as direct demonstration of new techniques.
WWW Applications - The lab has completed several web design projects and uses these projects to help students gain both web design and web technology experience. The lab is currently redesigning the website for the department of Computer Science to provide greater utility to the user community. This redesign involves both a re-structuring of the information and re-designing the look and feel of the site. This project has currently passed through two iterations of visual design review and is being implemented using a combination of HTML, PHP and MySQL.
The lab has developed several web applications ranging from trouble reporting systems for the internal use of the computational science group to an on-line survey system developed for the National Science Foundation. Current projects include the development of a problem solving environment for the PARINT system and a Computer Science website update system. On the right is a screen shot of the trouble reporting system for the computational science group.
Document Management - Document management systems most often consist
of a title + contents display where the user is shown a set of
document titles or brief descriptions and clicks on the title to
display the full document. We have developed three prototype
interfaces which all use thumbnail representations of text as document
surrogates. The most current of the interfaces is geared toward
keyword searching text documents. It attempts to focus the user on
keyword distribution within documents by using thumbnail
representations with highlighted areas of interest as document
surrogates and multi scale document representations showing keywords in
context. The image on the left shows an interface for browsing
documents which allows direct paging and extraction from thumbnail
images.