MODELLING DATA WAREHOUSE DIFFUSION USING FUZZY COGNITIVE MAPS - A COMPARISON WITH THE SYSTEM DYNAMICS APPROACH M S Khan 1 , M Quaddus 2 , A Intrapairot 3 and A Chong 1 1 School of Information Technology, Murdoch University, Perth, WA 6150 E-mail: s.khan@murdoch.edu.au, alexchong_@hotmail.com 2 Graduate School of Business, Curtin University of Technology, GPO Box U 1987, Perth, WA 6845 E-mail:quaddusm@gsb.curtin.edu.au 3 Rajamangala Institute of Technology, Northern Campus, Chiang Mai 50300, Thailand E-mail:intrapairota@cbs.curtin.edu.au ABSTRACT Fuzzy cognitive maps (FCMs) have been used recently for representing and analysing complex systems evolving with time. Results of such analysis can be used for decision support. This paper describes the use of an FCM for analysing the diffusion process of a data warehouse in a bank. An introduction to FCMs is given, and the process of building the FCM for simulating the data warehouse diffusion scenario is described. Analysis results obtained are presented and compared with the corresponding results obtained using the system dynamics methodology for modelling complex systems. Keywords: Fuzzy cognitive maps, decision support, data warehouse diffusion, system dynamics INTRODUCTION The core task of a decision support system (DSS) is decision analysis. Real-life problems are mostly unstructured in nature, which makes it difficult to apply algorithmic methods based on mathematical models to the process of decision analysis. Various analytical methods such as decision trees (Render & Stair, 1988), influence diagrams (Howard & Matheson, 1984) and systems dynamics modelling (Coyle 1996) have been developed as decision analysis tools. During the 80s and 90s a new breed of systems known as intelligent DSS has been developed, which mainly incorporates artificial intelligence (AI) techniques of knowledge representation and rule-based inferencing. More recently, neural networks, fuzzy logic and hybrid approaches have been tried. Cognitive maps (Axelrod, 1976), (Eden, 1990) and more recently fuzzy cognitive maps (FCM) have emerged as alternative tools for representing and studying the behaviour of systems. As yet, there have been few attempts at applying FCMs as modelling tools for simulating systems evolving with time, particularly in business applications. Data warehouse diffusion in an organisation may be viewed as a time-dependent process where a number of variables interact in a complex manner involving feedback. This paper describes an investigation into the applicability of FCMs for simulating data warehouse diffusion. In order to evaluate the performance of the FCM developed for this purpose, results obtained have been compared with those produced by the more well-established systems dynamics modelling tool. This paper first gives a brief introduction to the origin, evolution and application of the fuzzy cognitive map (FCM) as a new tool for decision support. A more detailed account may be found in (Khan et al. 2000). The next section outlines the significance of data warehouse diffusion as a means for benefiting from the use of information technology in an organisation. This is followed by a description of the FCM developed for modelling data warehouse diffusion. Next, an analysis of the FCM is given and a comparison made with simulation results obtained using the system dynamics approach. Finally, the issues of reliability and validity of the FCM as modelling tool is brought up before some concluding observations.