Business process instance model

Abstract

Business process management has grown into a mature discipline supported by a large number of commercial and open source products, collectively referred to as Business Process Management (BPM) systems. BPM systems store the process instance information in a physical storage known as Process Instance Repository. In an organisation several BPMS products can co-exist and work alongside each other. Each one of these BPM tools has its own definition of process instances, creating a heterogeneous environment. This reduces interoperability between business process management systems and increases the effort involved in analysing the data. In this thesis, we propose a common model for business process instances, named Business Process Instance Model (BPIM), which provides a holistic view of business process instances generated from multiple systems. BPIM consists of visual notations and their meta data schema. It captures three dimensions of process instances: process execution paths, instance data provenance and meta data. BPIM aims to provide an abstract layer between the process instance repository and BPM engines, leading to common understanding of business process instances.