Ulster University Logo

Ulster Institutional Repository

A Hybrid Knowledge Based System for Strategic Purchasing

Biomedical Sciences Research Institute Computer Science Research Institute Environmental Sciences Research Institute Nanotechnology & Advanced Materials Research Institute

McIvor, Ronan, Mulvenna, Maurice and Humphreys, Paul (1997) A Hybrid Knowledge Based System for Strategic Purchasing. International Journal of Expert Systems with Applications, 12 (4). pp. 497-512. [Journal article]

[img]PDF
Indefinitely restricted to Repository staff only.

2457Kb

Abstract

The aim of this paper is to show how knowledge-based systems technology can assist in the area of strategic purchasing. The authors discuss a knowledge-based system (KBS) designed to help companies in the make or buy decision, which is arguably the most fundamental component of manufacturing strategy. As part of the development process, procurement managers in ten multi-national companies were interviewed to determine current make or buy practice and elicit opinions on how the decision-making process could be enhanced. The general consensus was that a formal structure was needed along with computerized support at various stages in the process. The model was developed conceptually from the analysis of these interviews with procurement managers and also through a thorough literature survey of the area. The next stage was to computerize the most important components of the system to enable feedback from procurement managers in two of the multi-nationals first interviewed. Within the description of this KBS there is specific focus on the issues involved in the application of case-based reasoning (CBR) techniques and Multi-Attribute Analysis (MAA) to the automation of the make or buy decision. The development of this system is intended to illustrate that a case-based system should be capable of providing sound solutions utilizing relatively small case libraries, while avoiding a large rule base and long rule chains necessary if rule-based reasoning was used exclusively.

Item Type:Journal article
Faculties and Schools:Faculty of Computing & Engineering
Faculty of Computing & Engineering > School of Computing and Mathematics
ID Code:11616
Deposited By:Professor Maurice Mulvenna
Deposited On:25 Mar 2010 16:43
Last Modified:20 Mar 2014 11:59

Repository Staff Only: item control page