An Analytic Model of Success for Information Technology Decision Making

Abstract

Developing an information technology (IT) system to meet organizational needs is complicated, is often very extensive, takes a long time to realize, and is almost always costlier and more difficult than originally planned. To help with this complexity, many businesses use the Information Technology Infrastructure Library (ITIL) to guide the design, procurement, and operation of their IT systems. The ITIL is intended to optimally synchronize IT departments to function in accordance with the needs of the business. To further assist managers in monitoring the progress of their IT programs we developed a Bayesian Network stochastic model, the IT Decision Management System (ITDMS), to simulate the program's evidence observations, complex interrelationships, and the dynamic/temporal relationships. Based on the Defense Business Systems Acquisition Probability of Success (DAPS) Model, a technical framework developed at George Mason University, the model aligns the sub-process of each ITIL phase in a Bayesian structure that allows a decision maker to assess program performance in specific subject matter knowledge areas and the overall likelihood of program success by considering both data and temporal uncertainty. The key difference between DAPS and ITDMS is the explicit incorporation of the utility and decision factors in the Bayesian influence diagram model.