Implementing SOA Through Linked Data

Linked Data is a concise technical approach that can be leveraged to implement SOA. Some aspects are known from the REST architecture: every thing is a resource; URI design over service design; content negotiation. But the core feature of Linked Data is the linkage of datasets. Imagine an invoice directly linked to products and a customer by URI reference. As each dataset is rendered in RDF triples, everything can be accessed regardless to the respective schemas. If you want to know more, you may discover the schema statements on-the-fly (yet another URI reference in the data statement). Knowing the schema, you may post SPARQL queries and updates to federated endpoints. This makes perfect sense, for example, in a distributed masterdata scenario.

Would need to agree on some enterprise ontology before you start? Definitely not. Start with some simple vocabularies derived from your relational data model as it is. This may evolve to a more expressive ontology later, but „keep it simple” is the silver bullet!

Datum

06.10.2010

Uhrzeit

16:30 - 17:15

Konferenz / Veranstaltung

SOA & Cloud Symposium

Ort

Berlin