A metadata curation framework for data ecosystems

Abstract

A Data Ecosystem can be defined as a complex socio-technical network that enables collaboration between autonomous actors in order to explore data. Such ecosystems provide an environment for creating, managing and sustaining data sharing initiatives. While Data Ecosystems are thus arguably gaining importance, several ecosystems are not sustainable and consequently the effort spent by their actors end up not being properly used or forgotten. A comprehensive and meaningful description of all Data Ecosystem resources is needed. The increasing recognition of metadata as an essential asset had motivated an increased demand for metadata curation solutions. Metadata curation covers the creation and harvesting of metadata, appraisal and selection of metadata, quality assurance, preservation of metadata and other lifecycle stages, and also involves a number of IT systems. While important, in general, the current initiatives on metadata curation are a confusing mixture of activities, standards, terms and vocabularies, methods and tools. Referential guidelines would provide a basis to choose standard terms and definitions, processes and practices, roles and deliverables for metadata curation practitioners. In this context, this thesis aims to propose a framework, called Louvre, which offers a wide range of processes for aiding to curate metadata in Data Ecosystems. Each process describes a coherent set of engineering and management activities related to metadata curation. The Louvre structure is flexible and may be adapted to the needs of the actors interested in curating Data Ecosystem metadata. In this sense, processes are organized in functional dimensions, enabling modularization of the framework. Louvre also provides a set of best practices aligned with principles of agile and open collaboration for managing curation work through the collaborative effort of self-organizing actors. Finally, the framework is based on state-of-the-art in the area. This research also contributes to the Data Ecosystem area, by mapping the state-of-the-art of Data Ecosystems. In addition, it contributes also to the understanding of several issues related to the Data Ecosystems creation and maintenance. Also noteworthy is the definition, formalization and modelling of essential constructs related to Data Ecosystems.