The project support a "Open Data as a Service" infrastructure, making use of non-relational databases in order to better represent, inside the data model, the intrinsic linkability of information. The approach adopts Neo4j, thus natively conceiving data as a fully fledged Linked Data model using are Nodes and Relationships. This results in maintaining a crosscutting semantic within the data, which may evolve over time in terms of structures and relationships. This allows to easily manage data schemes which are closer to the end-users goals --- namely the terminal use cases to which the data will be applied to. We also allow data to be automatically structured in W3C compliant formats. Finally we promote data to be possibly enriched, through the use of third party APIs. Common use cases concern data to be analysed through NPL techniques, georeferenced or linked to mainstream data sets as DBPedia, etc.
The system is based on a plain Spring Data + Neo4J back-end. It is also characterized by a series of connectors through the external modules: input "gates", allows to feed the system starting from local files as well as data streams and APIs. Accordingly, it also provides a set of configurable output "gates", which allows data to be automatically retrieved, through the use of REST based APIs.
See here for additional details.