Data ecosystems reference architecture based on data mesh & data fabric

Abstract

Digital transformation requires rapid and profound changes to take advantage of technologies and data in order to make decision-making more effective with agility and self-sustainability. The complexity of data in the modern era and the silos that are generated at big scale drive the emergence of new data management models and architectures that focus on the intrinsic characteristics of digital ecosystems, characterized by the strong interrelationships of various actors through along the value chain, the platforms as the basis for interoperating with each other and the co-evolution of data products that emanate from increasingly heterogeneous sources. This article proposes the design of a reference architecture for data ecosystems based on the data architectures that are best supporting data management in this complex scenario: Data Mesh and Data Fabric, and with the use of knowledge graphs for the integration. As a method, an analysis of the most recent literature on data management and architectures is used to extract the principles and architectural components that are used in the design of such a reference architecture. An abstract representation of the reference architecture of data ecosystems is obtained, whose operational model is theoretically verified. It is the starting point for future research that will be directed towards its implementation in real use cases and organizational modeling related to the roles of the actors involved in the ecosystem reflected in the architecture itself.