Integrated Access to Multidisciplinary Data Through Semantically Interoperable Services in a Metadata-Driven Platform for Solid Earth Science.

The ability to use data produced by different sources (social networks, governments, weather sensors etc.) is widely recognized as a key to capitalize the value of data. In the scientific field, such usage may incredibly boost the innovation and foster new discoveries. However, one of the main hurdles is currently represented by the difficulties in achieving the required interoperability to provide integrated access to multi-disciplinary data. The current work presents a metadata-driven approach that uses in a combined way metadata, semantics, and services as key components for providing integrated access to heterogeneous data sources. The integration occurs within a central data integration system, which is driven by a rich metadata catalogue and that can present the data provided by the different data sources in a harmonised way to the end user, by means of RESTful APIs. A real application demonstrating metadata-driven semantic and service interoperability for achieving homogeneous access to multi-disciplinary heterogeneous data sources is illustrated in the case of EPOS, a Research Infrastructure for Solid Earth Science. The advantages in terms of ease of maintenance, of flexibility in plugging different standard without perturbating communities’ long-lasting technical practices, and of ability to track provenance are discussed. Future work for providing open-source implementation of a system built following the proposed approach is also envisaged.