Laurent Lefort
| Organization | CSIRO ICT Centre |
|---|---|
| Organization type | Research Institution |
| Country | Australia |
I am an ontologist in the team working on the Semantic Integration of web-based resources such as databases, web services and sensors networks. I develop domain models with ontology languages such as the Web Ontology Language (OWL) to support data integration in service-oriented and event-driven architectures.
I am also the co-chair of the Semantic Sensor Network Incubator activity at W3C and work on Environmental Sensing ontologies in collaboration with other CSIRO research teams. Finally, I am the deputy manager of the Australian W3C Office which is hosted by CSIRO ICT Centre since 2005 and one of the co-organisers of the Canberra Semantic Web Meetup group.
This member participated in the following Forums
Forum Forum: "Building the CIARD Framework for Data and Information Sharing" April, 2011
Question 3: What are the emerging tools, standards and infrastructures?
Data citation
One way to encourage the sharing of data is to develop the practice of data citation.
Here are two useful background documents, one from the Australian National Data Service (ANDS) and the other from the
ANDS and the other from Gen2Phen, an EU project focusing on Health and Life science research data.
- Data Citation Awareness http://ands.org.au/guides/data-citation-awareness.html
- D9.3 Draft Report on Incentives and Rewards in the Field of Biomedical Research Databases http://www.gen2phen.org/system/files/private/D9.3%20Draft%20Report%20on…
Standards and tools to transform data from SQL to RDF
I recommend to RDF beginners to start with tools which implements a Direct Mapping from a database. Direct Mapping is defined as a mapping that mirrors the database schema in RDF with a minimal effort required to implement it. There have also been efforts to let users annotate the SQL code to provide the same capability e.g the work done by the FlyWeb project
- My first mapping from RDB to RDF using a direct mapping http://ivan-herman.name/2010/11/19/my-first-mapping-from-direct-mapping/
- Future of FlyWeb work on Chado OWL ontology and RDF mapping http://generic-model-organism-system-database.450254.n5.nabble.com/Futu…
These tools are not only interesting for users willing to transform an existing database. MOLGENIS http://sourceforge.net/projects/molgenis/ is a tool to allow users to create their own TAB-delimited format to record science data and then move it into RDF (using D2RQ) to allow other apps to process it. The starting point is a couple of XML files (one for the data model and one for the UI) with a simple syntax (there is a utility to extract the data model one from an existing database). Out of these original "models", the MOLGENIS project aims to derive a range of tools including a R API and RDF access using D2RQ in a comparable way to what has been done in the FlyWeb project.
LOD and ontology
The gap between "lightweight semantics" like LOD and ontology-based approach is much smaller than it used to be. The Datatype Reasoning capabilities enabled by the OWL2 standard http://www.w3.org/TR/owl2-overview/ and the new features provided by ontology engineering tools like SPARQL-DL http://www.w3.org/2001/sw/wiki/SPARQL-DL can help LOD users to exploit ontology content even if it is mixed with numerical data.
Graph databases (NoSQL)
Finally, Graph databases may also have a role to play in future Linked Open Data infrastructures because there are new (and also old) products now fighting over a market niche which is roughly half way between traditional databases and triple stores.
Sandro Hawke Toward Standards for NoSQL NoSQL Live … from Boston March 11, 2010 http://www.w3.org/2010/Talks/0311-nosql/talk.pdf