Detecting and explaining business exceptions for risk assessment

Systematic risk analysis can be based on causal analysis of business exceptions. In this paper we describe the concepts of automatic analysis for the exceptional patterns which are hidden in a large set of business data. These exceptions are interesting to be investigated further for their causes and explanations. The analysis process is driven by diagnostic drill-down operations following the equations of the information structure in which the data are organised. Using business intelligence, the analysis method can generate explanations supported by the data.