This project proves that the semantic technology stack and related tooling allow to do data integration in a fast and agile way. Several technologies have been utilised: Ontology-Based Database Access, federated SPARQL, federation middleware. The solution is using the RDF Data Cube vocabulary for capturing the emission observations done. The additional 5 star LOD publishing was easily achieved at a minimal cost.
The ‘Omgeving’ Department is the environmental administration of the government of Flanders. It is responsible for preparing, following up and evaluating the Flemish environmental policy. In Belgium, companies that want to emit polluting substances in the air or water must have an environmental permit. Some of them are also obliged to report annually about the emissions of the previous year. In fact, data regarding the amount of substances per location have been collected since 2004.
The idea was to integrate these data with all kinds of other data sources (internal and external) to offer: • the general public an application showing the emitted substances over the years in the area that they live • companies the ability to benchmark themselves against other, similar, organisations • public servants analytics dashboards to gain insights to steer the policy making. The ultimate aim was that by integrating the data (research) questions can be addressed not being able to be answered on the separate data silos.
The solution preferably needed to be based on open source software or using open source libraries.
The project serves also as a pilot site of the OpenGovIntelligence project[
Dataintegration. The actual status is that more than 10 data silos have been integrated. These datasets contain data managed by the department itself (archival systems, RDBMS) and datasets published by other departments (Flemish addresses database, Belgian company register) together with well-known classification systems such as NACE (economical activities) and NIS (administrative geographical entities).
The solution takes a hybrid approach. For some datasets, the data are transformed
via dedicated ETL processes into triples and these are loaded in a triple store. Other
datasets, mainly from existing relational databases, are virtualized via OBDA[
General Public Dedicated Application. An application has been build showing the
emitted substances over the years in the area that they live. This is a traditional web
application using Polymer [
Business Intelligence. A connector has been developed for exploratory [
The approach used has clearly proven that semantic technologies allow to do data integration in a fast and agile way. Answers can now be given to questions which involve data from multiple datasets. The fact that 5 star LOD publishing is only one step away is a nice add-on. The system is used internally for the moment and will be become open to the public early November.