Knowledge Driven Behavioural Analysis in Process Intelligence

In this paper we illustrate how the knowledge driven Behaviour Analysis, which has been used in the KITE.it process management framework, can support the evolution of analytics from descriptive to predictive. We describe how the methodology uses an iterative three-step process: first the descriptive knowledge is collected, querying the knowledge base, then the prescriptive and predictive knowledge phases allow us to evaluate business rules and objectives, extract unexpected business patterns, and screen exceptions. The procedure is iterative since this novel knowledge drives the definition of new descriptive analytics that can be combined with business rules and objectives to increase our level of knowledge on the combination between process behaviour and contextual information.