Modularizing Monitoring Rules in Business Processes Models

Abstract

Business process management systems contain monitoring, measurement and control (MMC) specifications to enable the identification of problems and solutions to improve business processes. Business processes facilitate the integration of human and technological resources in an organization, according to a set of activities that fulfil a policy goal [1].

Currently, several business-process monitoring, measurement and control solutions are available [2]. However, MMC specifications typically are implicitly encoded in the low-level implementation of the workflow system. This results in tangled and scattered MMC knowledge in the underlying process code. It is clear that this decreases the maintainability and reusability of the MMC specifications since they are not specified in a modular fashion. Furthermore, specific knowledge about the overall system implementation is required if the MMC requirements evolve. Due to the entanglement with the low-level implementation, this requires a level of expertise that is normally available to technical developers. This is unfortunate since business experts typically express the MMC requirements at a high level and in terms of the business domain (as opposed to the technical implementation). The task becomes even more complicated because the MMC specifications involve data that is not readily available in one location of the process code. In addition the majority of existing approaches focus on describing the MMC specification in terms of the process execution instead of the data flow of the process. Consequently, when existing MMC specifications need to be adapted, the different pieces of process code must be manually localized after which adaptations will occur at several places.