| Fraude and fraudulent business practices are of all times. Legislators, regulators and financial authorities try to create legal frameworks and procedures for early detection and prevention, but they have failed consistently during the last 10-15 years. This proposal will create a unique cooperation between experts in Dutch auditing theory and software engineering researchers specialized in software analysis and domain-specific languages. The grand vision of this proposal is to create auditing services that can perform real-time monitoring of companies in order to fundamentally increase the reliability and transparency of financial systems and supporting IT systems. We start with the design of a meta-model for auditing and create the concepts and techniques to describe, analyze, and maintain specific company models and to automatically generate auditing services from them. These services include intake (establishing a company model based on the mining of available data and source code) and monitoring and checking (comparing actual data with the company's model). Key elements of our approach are the use of software analysis to extract facts from financial reports and source code of IT systems, to link the software with actions in the company's model, and to address global issues on maintenance and evolution of these models. The increased trust that is created by these advanced auditing services is in itself an enabler for software services in general. |
