Files in this item



application/pdf9416392.pdf (11MB)Restricted to U of Illinois
(no description provided)PDF


Title:Development of procedural expertise to support multiattribute spatial decision-making
Author(s):Lee, Insung
Doctoral Committee Chair(s):Hopkins, Lewis D.
Department / Program:Transportation
Urban and Regional Planning
Urban and Regional Planning
Degree Granting Institution:University of Illinois at Urbana-Champaign
Urban and Regional Planning
Abstract:This study focuses on the development of process oriented support for planning support systems. To support exploration of ill-structured planning problems, planning support systems should be incorporated within planning processes, decision-oriented rather than just analysis-oriented, and able to provide effective procedural supports.
To attain these claims about good planning support systems, a DSS Generator for multi-attribute judgment was developed. Complementary, synthetic use of four judgment methods--MAUT, AHP, CEA and ELECTRE--and heuristic rules was explored in constructing the judgment module. Procedural support strategies for ill-structured judgment methods were developed by Monte-Carlo simulation and implemented in the module. To demonstrate the effectiveness of this module, it was integrated into a DSS, PEGASUS, that also supports the design phase. A specific planning situation--a bus routing problem--was identified, and a module for alternative generation was constructed.
Evaluation of the system showed that the system fulfilled what it was designed to do. Observations of use patterns of the system showed that access to multiple judgment methods and alternative generation methods allowed users to explore problems constructively and iteratively. The procedural supports were accepted by users on most occasions and allowed them to use the judgment methods spending less effort on procedural judgments. The system thus proved by example that some procedural expertise can be identified and that this expertise can be brought to bear in an operational system.
A number of implications for future research were raised through this study. Suggested directions for future research include the application of PEGASUS to other planning domains, extension of procedural support for other phases of the decision making process, and exploration of more active use of value information in the design phase. Some perceptual observations--the different learning curves of the judgment methods and the complementary roles of multiple judgment methods--also deserve further exploration with an experimental design.
Issue Date:1994
Rights Information:Copyright 1994 Lee, Insung
Date Available in IDEALS:2011-05-07
Identifier in Online Catalog:AAI9416392
OCLC Identifier:(UMI)AAI9416392

This item appears in the following Collection(s)

Item Statistics