Supporting compliance verification for collaborative business processes.

Abstract

Collaborative business processes are the current trend of business processes supported by the advances in technology like the Internet and collaborative networks. Enterprises no longer do business in isolation. The customer demands are always changing and becoming sophisticated with dynamic requirements and the shortening period in which they must be met. Collaborative business processes must conform with not only customer demands but also with laws, standards, best practice and regulations. These impose constraints on the business process that must be satisfied otherwise they attract criminal charges or financial fines. Corporate scandals for companies like Enron, World- com, Societe General etc. were a result of non-compliance. This attracted regulations like the Sarbanese Oxley Act, Basel III, Anti money laundering act among others with articles guiding operational practice. However, non compliance is still observed especially among SMEs that do not possess the skilled man power or the funding to acquire automated compliance solutions. In this thesis, we sought to support non-expert end users through a compliance management approach that can guide the specification and verification of compliance for collaborative business process with a range of policy and regulatory requirements. Collaborative business processes differ from traditional business processes. They are characterised by specific attributes that present unique verification requirements that cannot be automatically addressed by existing verification approaches. To achieve the intended goal, design science research method was employed to develop a mechanism to elicit requirements from different sources, translate them into formal constraints based on formal semantics, and a set of algorithms were composed to support compliance verification. The algorithms provide meaningful and easy to understand feedback to the end user about the compliance or violation of the collaborative business process. Due to the fact that policies and regulations change often, we adopted simulation analysis as a technique to assess and analyse the impact of such changes to the business process before actual implementation. The thesis artifacts are evaluated based on known information systems model evaluation methods following the design science recommended steps and the Method Evaluation model (MEM). We also validate and evaluate the compliance algorithms using a different industrial use case (the car insurance trading business process) from the case used in their design (the pick and pack business process). Further more, the performance of the algorithms is evaluated based on their computation complexity.