End-User Visualization and Manipulation of Distributed Aggregate Data

T. Paul McCartney, Kenneth J. Goldman
Aggregate visualization and manipulation enables the viewing and interaction of dynamically changing data sets in a graphically meaningful way. However, off-the-shelf applications typically provide only limited ways to view static aggregates and generally do not support manipulation of aggregate data through the resulting visualization. To be fully dynamic, an aggregate visualization should be customizable to suit the individual's needs and should allow end-users to modify the data through direct manipulation. This paper describes a software system that empowers end-users to create interactive aggregate visualizations through a visual language interface. Included are mechanisms for specifying how aggregate data is processed from multiple sources of a distributed application, providing functionality similar to project, select, join, and cross product of relational databases. This approach gives end-users the power to create customized, interactive visualizations of dynamically changing aggregate data without the need for textual programming.

KEYWORDS: aggregate data, constraints, direct manipulation, distributed computing, matching, multi-way constraints, user interface management system, visualization

Available as postscript (.ps) or compressed postscript (.ps.gz).


Washington University Department of Computer Science WUCS-97-48, December 1997.

Submitted to Journal of Visual Languages and Computing.


Prepared by T. Paul McCartney (paul@cs.wustl.edu)
Washington University Department of Computer Science