Assignment 2: RDF

RDF’s role for metadata processing

• RDF provides a model and syntax for representing metadata.
• More basically it provides a model and syntax to make statements about properties of resources. Resources can be properties of other resources.
• Relationships between resources and class hierarchies can be defined using this syntax.
• As all information about a resource is modeled in terms of properties, which can have certain values, it is very easy to gain/define any knowledge about a resource.
  On the other hand redundant information has to be supplied. E.g. all resources, which are members of a certain class have to contain a property describing their type.
• RDF can be used to describe ontologies.
• RDF can be used to describe properties of resources.
• The syntax is built on top of the existing XML syntax. This helps integrating RDF in existing file formats, which are used on the web.

Examples

1. Use of a schema

The following schema expresses some relationships between different kinds of parrots.

<rdf:RDF xml:lang="en"
 xmlns:rdf="http://www.w3.org/1999/02/22-rdf-syntax-ns#"
 xmlns:rdfs="http://www.w3.org/2000/01/rdf-schema#">
 
<!-- Note: this RDF schema would typically be used in RDF instance data 
 by referencing it with an XML namespace declaration, for example
 xmlns:xyz=" http://www.arches.uga.edu/~bernie42/parrots.rdf". This allows
 us to use abbreviations such as xyz:parrot to refer
 unambiguously to the RDF class 'parrot'. -->
 
<rdf:Description ID="parrot">
 <rdf:type resource="http://www.w3.org/2000/01/rdf-schema#Class"/>
 <rdfs:subClassOf
 rdf:resource="http://www.w3.org/2000/01/rdf-schema#Resource"/>
</rdf:Description>
 
<rdf:Description ID="Cockatoo">
 <rdf:type resource="http://www.w3.org/2000/01/rdf-schema#Class"/>
 <rdfs:subClassOf rdf:resource="#parrot"/>
</rdf:Description>
 
<rdf:Description ID="Lory">
 <rdf:type resource="http://www.w3.org/2000/01/rdf-schema#Class"/>
 <rdfs:subClassOf rdf:resource="#parrot"/>
</rdf:Description>
 
<rdf:Description ID="CardinalLory">
 <rdf:type resource="http://www.w3.org/2000/01/rdf-schema#Class"/>
 <rdfs:subClassOf rdf:resource="#Lory"/>
</rdf:Description>
 
</rdf:RDF>

If someone wanted to know something about his Cardinal Lory a search engine could use the above hierarchy and conclude that resources, which contain information about Lories or about parrots could contain information about Cardinal Lories. Therefore the search takes more relevant resources into account.

2. Information about resource

Lets assume I want to know who worked only on the latest versions of a certain Linux distribution. I want to know about any new programmers. If the website of that company would supply metadata about the different versions of their distribution using the Dublin Core a search engine could calculate the information I am looking for.

The search engine could look at Creator elements of different versions and at the Date and Version elements. Those people who are only mentioned in the version from the latest Date are those that I am looking for.

RDF annotated resource

http://www.aifb.uni-karlsruhe.de/~sst/. There are a lot of examples about how to use RDF, e.g. here.

Tool: Protégé-2000^TM

This tool has been developed by Standford Medical Informatics. It allows the user to construct a domain ontology and to enter domain knowledge.

It can import and export ontologies in the following formats: Text file, JDBC Database, RDF Schema. Further Protégé-2000 can create a javadoc-style html-file for documentation. The terminology used in the user interface seems to vary from RDF terminology. E.g., instead of “range” the term “slot” is used.

The tool is written in Java and comes with a sophisticated installer. For more information and download see http://protege.stanford.edu/index.shtml.

Research projects

1. Web Resource Application Framework

The goal of this project is to constructing an application, which is completely integrated in the RDF data model.

All data, the user interface is described in RDF. The user interface could be automatically adjusted to a user profile, or to the data it is actually handling.

The program code is fit into the model as RDF Literals. This would make it easy to download and integrate new code.

For more information see http://wraf.org/.

2. FRODO (A Framework for Distributed Organizational Memories)

This project is about methods and tools for building and maintaining distributed Organizational Memories in a real-world enterprise environment.

To construct those tools techniques, which are known

from building agents for workflow enactment and information access,

from ontology acquisition from texts and user interaction and

from document analysis and understanding

are used.

These tools include the rdf2java tool, which can create Java objects directly from RDF files,

a tool for visualization of ontologies represented as RDF schemas,

a logic programming language, which is based on RDF and

a tool for editing and classifying OIL ontologies with Protégé-2000.

For more information see http://www.dfki.uni-kl.de/frodo/index.html.