» Introduction.

» What Does RDF Stand For ?

• Resource
• Descriptor
• Framework

» What Is RDF ?

• RDF provides a formal way of defining 'meaning' for meta-data.
• RDF can be used to 'tag' items with descriptions that are machine interpretable.
• RDF is one of the key building blocks of the Semantic Web.

» Where Is RDF Used Today ?

• RDF can be used to embed information about pictures in image files.
• RDF can be used to add machine interpretable 'about' information to web pages.
• RDF 'triples' are building blocks used to build ontologies for inference engines.

» Why are We Interested in RDF ?

• The management of data is a fundamental problem in large systems architectures.
• Different systems use different data models and schemas to store information.
• Meta-Data helps systems to interpret and process their data.
• RDF can enable one system to 'understand' another's meta-data, and hence its data.

» Why Does this Matter to Us ? - The 'Many Interfaces' Problem ...

• In large-scale systems architectures there are often many component systems.
• A huge amount of effort is spent in making systems 'talk to one another'.
• We can use meta-data ontologies to help use 'glue' diverse systems' data models together.

» Why Does this Matter to Us (Again) ?

• In quantitative financial risk management there are often a large number of analytical results.
• The 'meaning' of such results is often deeply embedded in the 'business logic' of systems.
• This makes such systems 'closed', difficult to inter-operate, and fragile to change.
• We can attach machine interpretable meta-descriptors to analytics results, and solve this.

» The Structure of RDF.

• RDF is made up of 'triples'.
• Each triple represents a 'fact' or 'statement'.
• Each triple contains :
  - A subject.
  - A predicate.
  - An object.

» Representing RDF Triples.

• An RDF triple can be represented as a directed graph.
• {Subject} --- predicate ---> {Object}
• Many inter-related RDF triples can be represented as a single, more complex, RDF graph.

» The Serialization of RDF.

• The official form is RDF/XML, for example:

    <rdf:RDF>

      <rdf:Description rdf:about="subject">

        <predicate>object-value</predicate>

      </rdf:Description>

    </rdf:RDF>
          

» RDF, RDF-S and OWL.

• RDF
  -  Resource Descriptor Framework.
  -  A mechanism for defining fact statements as 'triples'.
• RDF-S
  -  RDF-Schema.
  -  A set of standard RDF entities for building vocabularies.
• OWL
  -  Web Ontology Language (re-ordered!).
  -  Highly expressive Ontology language - W3C sponsored.

» Why Have I not Heard of this Before ?

  - Some aspects of this are genuinely difficult:

• At the moment, the techniques and tools required tend only to be accessible to experts.
• This kind of technology often works 'behind the scenes' as an 'enabler'.
• These technologies have not yet found their way to far into 'Consultancy Babble'.

» Applications In Data Management

• US agencies such as 'Homeland Security' use these methods to tie together the vast amount of information that they have spread over many systems and databases.
• Many other institutions face similar data and systems dilemmas.
• We can define ontologies, and use these as data 'meta-meta-models'.

» Conclusions.

• RDF, and the 'ontologies' family has the potential to be an extremely powerful technology.
• There are significant barriers to entry.
• The 'semantic web' is in its infancy, as it evolves these methods will be more widely adopted.
• These technologies will be the basis for a new generation of data management solutions.