A Bayesian belief network for IT implementation decision support

Bayesian Belief Networks (BBNs) are graphical models that provide a compact and simple representation of probabilistic data. BBNs depict the relationships among several variables and include conditional probability distributions that make probabilistic statements about those variables. This paper demonstrates how to create a BBN from real-world data on Information Technology implementations. The paper also displays the resulting BBN and describes how it can be incorporated into a DSS to support ‘what-if’ analyses about Information Technology implementations. The paper combines techniques originating from artificial intelligence, statistics, and computer-based decision making.