Information Society uses information and generates data: companies store their transactions with costumers and suppliers, hospitals keep track of their patients’ histories, public institutions record the behavior of citizens with respect to taxes, justice, health-care. All this information is stored in large and fast-growing databases. These databases represent a management challenge and an invaluable wealth: the challenge of effectively navigating through this sea of information and the wealth that could derive from the exploitation of these data to enhance planning, prediction and decision making. The newborn field of research known as Knowledge Discovery in Databases (KDD) is meant to meet these challenges by developing methods and techniques able to extract useful and reusable knowledge from real-world databases.

Starting from June 10th, 1997 the Knowledge Media Institute will distribute a new Knowledge Discovery program called Bayesian Knowledge Discoverer (BKD). BKD has been developed within a collaborative project between KMi and the Department of Actuarial Science and Statistics of City University. BKD is a computer program able to extract Bayesian Networks – also known as Causal Probabilistic Networks – from (possibly incomplete) databases, using a novel estimation method devised by researchers of KMi and City University. The methodology underlying BKD blends together statistical theories due to an Eighteen Century priest – the rev. Thomas Bayes, in the picture – with the most advanced Artificial Intelligence techniques for machine learning and automated reasoning under uncertainty.