Enhancing the ENVISION Interface
for Digital Libraries

Jun Wang, Abhishek Agrawal,
Anil Bazaz, and Supriya Angle
Department of Computer Science
Virginia Tech
Blacksburg, VA 24061
+1 540 231 3615
{juwang2, aagrawal, abazaz,
sangle}@csgrad.cs.vt.edu
Edward A. Fox
Chris North
Department of Computer Science
Virginia Tech
Blacksburg, VA 24061
+1 540 231 6931
{fox, north}@cs.vt.edu

ABSTRACT
To enhance the ENVISION interface and facilitate user
interaction, various techniques were considered for better
rendering of search results with improved scalability. In this
paper we discuss the challenges we encountered and our
solutions to those problems.
Categories and Subject Descriptors
H.5.2 [User Interfaces]: Graphical user interfaces (GUI),
Screen design, Theory and methods.
General Terms
Algorithms, Design, Human Factors.
Keywords
Information Visualization, Digital Library, ENVISION,
Aggregation, Variable Width.
1. INTRODUCTION
The Envision interface helps visualize search results from the
MARIAN digital library system [1]. The old Envision interface
(Figure 1) featured a Query Screen (left), Result Visualization
Screen (right), and Result List Screen (bottom). Envision
presented each document in the result set graphically as an icon.
The user could select several documents on the Visualization
Screen and see corresponding details in the Result List Screen
[2].
Envision used a rigid matrix to display search results. As with
various other visualization tools, it provided no overview, thus
no context information to users. Scrollbars were needed for data
outside the viewable area. In working to enhance Envision, we
considered other efforts. Some approaches (e.g., ThemeScapes
[3]) are almost exactly opposite, displaying data in a completely
flexible manner. The data is not bound to any axis but is
arranged according to relationships between documents. We
now explore a middle ground with an overview as well as loose
matrix arrangement, with related documents mapped together, as
shown in Figure 2. The major changes considered are described
in Sections 2-4 below.





















2. AGGREGATION
With some visualization tools, users face the “keyhole” problem,
wherein they can only see a part of the entire visualization at a
time, with scrollbars used frequently [4]. In the old Envision
interface, with typical query results, much screen space was
empty. At the same time, few details were visible for large
clusters. In one case with 71 results, and a screen of 144 cells,
20 screen spaces were required to see all results. We use an
aggregation algorithm to get rid of scrollbars and visualize
everything in one screen space. Adjacent columns/rows are
aggregated together. Consequently, either the X or Y-axes
might have more than 1 attribute value per unit.





Permission to make digital or hard copies of all or part of this work for
personal or classroom use is granted without fee provided that copies
are not made or distributed for profit or commercial advantage and that
copies bear this notice and the full citation on the first page. To copy
otherwise, or republish, to post on servers or to redistribute to lists,
requires prior specific permission and/or a fee.
JCDL’02, July 13-17, 2002, Portland, Oregon, USA.
Copyright 2002 ACM 1-58113-513-0/02/0007…$5.00.

Figure 1. Old Envision interface
Figure 2. New Envision interface
275