Towards a Holistic Method for Business Process Analytics

In this paper, we propose a holistic approach aimed at combining business process modelling and data-driven business process improvement. The first step requires to develop a "precise" model of the processes of the organization using the UML. Precise means that all business entities involved in the process are determined as well as all the tasks composing the process executions, and all relevant data about them are modelled. Then, a model of the data space of the process will be derived taking into account also the quantitative aspects of each process (e.g. how many instances of the process will run each day?, how many possible instances of some business entities will be around or have ever been created?). In this way it is possible to conceive and design various analyses and improvements of the process based on its data, since all the aspects related to each business process have been explicitly modelled, and in a sufficiently formal way. We will introduce our approach using a small case study: the Buying process having as participants manufacturers, dealers, shippers, and payment systems.