Auditable Digital Data Market
A digital data marketplace (DDM) is an environment or platform enabling data sets to be exchanged, or listed, browsed and purchased. Typically, organizations or individuals participate for various benefits (e.g., commercial or scientific) of data publication, while some extend offerings to help other parties manage and analyze data, such as sharing of analysis algorithms and data-processing services. A DDM constructs a distributed computing platform allowing providers and consumers to exchange data assets and creates a trusted infrastructure for data processing with potential conditions. Data markets facilitate exchanging useful business data among enterprises, possibly generating great value by sharing and utilizing of such data. However, security and privacy concerns may arise, especially in cases where parties are generally competitors. The model of a DDM and the infrastructure must therefore be adequate in tackling such concerns: Consortium governance may be used to prevent asset exposure by ensuring all transactions are compliant to agreements between all parties.
Current project studies a new DDM model integrating automated auditing processes in collaboration with AMdEx field-lab. Firstly, an understanding and overview of recent models and requirements is obtained. Secondly, identification of logging and/or monitoring techniques is required that can be used to obtain the required data in the distributed setting of DDM implementations. Thirdly, methods for converting low-level system events to higher-level (legal) audit trails are to be explored, such as domain-specific languages for event processing. Lastly, based on eFLINT domain-specific language can be used to automate the assessment of audit trails against regulations, agreements, and resource-level conditions. The resulting pipeline should be generic in that it can support various use cases and data exchange applications. Additional challenges include applying the pipeline to automatically assess audit trails against multiple interpretations of the regulations, agreements, and conditions relevant to a case. A conceptual DDM model will be created as well as the realization of a testable demo model.