Date of Award


Degree Type


Degree Name

Doctor of Philosophy (PhD)

Graduate Group

Computer and Information Science

First Advisor

Norman Badler


Much recent research has been focused on user interfaces. A major advance in interface design is the User Interface Management System (UIMS), which mediates between the application and the user.

Our research has resulted in a conceptual framework for interaction which permits the design and implementation of a UIMS generator system. This system, called Graphical User Interface Development Environment or GUIDE, allows an interface designer to specify interactively the user interface for an application.

The major issues addressed by this methodology are making interfaces implementable, modifiable and flexible, allowing for user variability, making interfaces consistent and allowing for application diversity within a user community.

The underlying goal of GUIDE is that interface designers should be able to specify interfaces as broadly as is possible with a manually-coded system. The specific goals of GUIDE are:

  • The designer need not write any interface code.
  • Action routines are provided by the designer or application implementator which implement the actions or operations of the application system. Action routines may have parameters.
  • The designer is able to specify multiple control paths based on the state of the system and a profile of the user.
  • Inclusion of help and prompt messages is as easy as possible.
  • GUIDE's own interface may be generated with GUIDE.

GUIDE goes beyond previous efforts in UIMS design in the full parameter specification provided in the interface for application actions, in the ability to reference application global items in the interface, and in the pervasiveness of conditions throughout the system. A parser is built into GUIDE to parse conditions and provide type-checking.

The GUIDE framework describes interfaces in terms of three components:

  • what the user sees of the application world (user-defined pictures and user-defined picture classes)
  • what the user can do (tasks and tools)
  • what happens when the user does something (actions and decisions)

These three are combined to form contexts which describe the state of the interface at any time.