Publication of the Content of BGR's Information Systems to the Semantic Web

Just before the rise of the Web 2.0 Tim Berners-Lee, James Hendler, and Ora Larissa published an article on “The Semantic Web”, which is also known as “Web 3.0” or the “Web of Data”. It contains a general description of how to express the meaning of data and to present knowledge in the World Wide Web in order to make it more interoperable and machine-friendly. Several W3C standards on an appropriate data model (RDF), the declaration of vocabularies (RDFS and OWL), and a versatile query language (SPARQL) were up to follow. Since then it is possible to work with standardized formats and vocabularies, such that computer programs do not need to learn too much in order to understand the meaning of the data and to follow the linkages between them. Therefore, machines, software agents or computer programs should be able to search data on the WWW for a given purpose and present more concise and more complete results. At the Federal Institute for Geosciences and Natural Resources (BGR) a number of information systems have been reworked during the past three years. In this context, RDF/SPARQL interfaces will be added to these systems in order to increase their interoperability. This will be achieved by using a software applying W3C’s R2RML, the RDB to RDF Mapping Language. Problems we are facing are poorly normalized relational schemas, for which this kind of mapping is not well-defined (e. g. the database of the “Lithostratigraphic Dictionary”). Other problems not yet solved are the stable online publication of large amounts of data as in case of the “Geoscientific Collections” database with millions of entries. Some of these problems can be addressed by database normalization. Others will be solved by changing from transient (in-memory) to persistent Linked Data Views. The setup of RDF/SPARQL interfaces for BGR Information Systems is still in progress but seems to be a good preparation for data discovery by using standardised exchange formats and vocabularies and by resolving the linkages between the resources