Decision support system induced guidance for model formulation and solution

  • Authors:

  • Reza Barkhi;Erik Rolland;John Butler;Weiguo Fan

  • Affiliations:

  • Department of Accounting and Information Systems, Pamplin College of Business, Virginia Polytechnic Institute and State University, Blacksburg, VA;The A. Gary Anderson Graduate School of Management, University of California, Riverside, CA;Department of Accounting and MIS, Fisher College of Business, The Ohio State University, Columbus, OH;Department of Accounting and Information Systems, Pamplin College of Business, Virginia Polytechnic Institute and State University, Blacksburg, VA

  • Venue:
  • Decision Support Systems
  • Year:
  • 2005

Quantified Score

Hi-index 0.00

Visualization

Abstract

One of the critical functions of Decision Support System (DSS) is to provide system induced decision guidance for proper model formulation and solution. We show how to incorporate this type of system induced decision guidance into the design of the next generation of DSS. We suggest that a DSS should make decisions, or at least recommendations, regarding what models should be executed to solve problems most effectively and this information should be generated inductively and used deductively. This information then becomes the meta-model to induce the user to make appropriate choices. We provide an example that will illustrate how two specific problem characteristics, namely the tightness of constraints and the linearity of constraints, influence the solution quality and solution times for a specific class of test problems. We argue that a DSS should execute different formulations of the problem that lead to satisficing solutions guiding DSS users in finding the best approach to solve complex problems.