A Process Warehouse for Process Variants Analysis

Abstract

Process model variants are collections of similar process models evolved over time because of the adjustments that were made to a particular process in a given domain, e.g., order-to-cash or procure-to-pay process in reseller or procurement domain. These adjustments produce some variations between these process models that mainly should be identical but may differ slightly. Existing approaches related to data warehouse solutions suffer from adequately abstracting and consolidating all variants into one generic process model, to provide the possibility to distinguish and compare among different parts of different variants. This shortcoming affects decision making of business analysts for a specific process context. This paper addresses the above shortcoming by proposing a framework to analyse process variants. The framework consists of two original contributions: (i) a novel meta-model of processes as a generic data model to capture and consolidate process variants into a reference process model; (ii) a process warehouse model to perform typical online analytical processing operations on different variation parts thus providing support to decision-making through KPIs; The framework concepts were defined and validated using a real-life case study.