On warehouses, lakes, and spaces. The changing role of conceptual modeling for data integration

The role of conceptual models, their formalization and implementation as knowledge bases, and the related metadata and metamodel management, has continuously evolved since their inception in the late 1970s. In this paper, we trace this evolution from traditional database design, to data warehouse integration, to the recent data lake architectures. Concerning future developments, we argue that much of the research has perhaps focused too much on the design perspective of individual companies or strongly managed centralistic company networks, culminating in today's huge oligopolistic web players, and propose a vision of interacting data spaces which seems to offer more sovereignty of small and medium enterprises over their own data.