The Enterprise Data Marketplace : a platform for democratizing company data

Abstract

In the era of big data, multitudes of data are generated and collected in companies. This data contains a potential value that can be leveraged to gain new insights, e.g., for enhancing business models or reengineering industrial products. Extracting data value requires that this data is available for use. Yet, studies show that significant amounts of the data remain unused in companies. In this regard, data democratization initiatives, with the goal of empowering company employees to find, understand, access, use, and share data, are gaining in importance. Towards this end, data marketplaces are studied as metadata-based platforms to facilitate the exchange and provisioning of data and data-related services. However, data marketplaces are mainly investigated for the exchange of data and services between organizations or private individuals, i.e., as external data marketplaces. Little research focuses on the use of data marketplaces in the company-internal context, i.e., as an Enterprise Data Marketplace (EDMP). Topics of how the EDMP differs from an external marketplace, the scope of its offerings and functionality, challenges that arise in the company-internal context, or how the EDMP can be embedded in and leverage the existent company system and tool landscape have not been investigated in detail thus far. In this thesis, the Enterprise Data Marketplace is examined as a platform for democratizing data in companies, and in this context, the above-listed gaps are addressed. To this end, four research goals (RGs) are put forward: (RG1) the identification of the processes and challenges company employees face in finding, understanding, accessing, and sharing data in the enterprise without an EDMP; (RG2) the identification of the distinctive aspects of an EDMP; (RG3) establishing an architectural foundation for building an EDMP; and (RG4) the goal of leveraging existent metadata in the company tool and system landscape in the EDMP. The research goals are covered through nine research contributions. These entail the current data provider and data consumer journeys and challenges, an EDMP type distinction based on an EDMP definition, as well as a presentation of its distinctive characteristics, requirements, and challenges. An enterprise integration and platform architecture, together with an approach for leveraging existent metadata, yields the foundation for building an EDMP. The feasibility of the concepts put forward in this thesis is demonstrated through an EDMP prototype and an evaluation based on an experiment and qualitative assessment. The evaluation yields that the EDMP is well-suited for the effective realization of data democratization within companies and that it not only addresses several of the current issues data providers and consumers face but also increases the efficiency and reduces the complexity in accessing data. This thesis therefore introduces the EDMP as a platform for democratizing company data and lays the foundation for establishing the EDMP within a company.