Web 3.0: semantic addition to the searches
- Data has the value of meaning something. To describe this meaning in a global way it’s important a standard machine-readable format.
- On the big web, data is scattered under different labels, depending on the language (for example name, nom, nombre, etc.), the code of the center, etc. It’s important to merge this equal data when it refers to the same object, so we see a meaningful integration of data.
RDF (Resource Description Framework) is a semantic web tool. It describes and creates relations among resources on the web and interchanges this data. Its base are statements, like Oslo is a city, Oslo is in Norway and so on. The syntactical model is the triplet: (s p o), where s names the relationship between p and o. Note that the relationship must have a name, and URIs 1 (Uniform Resource Identifier) are universal tools for naming any piece of information (including properties or relationships). They ground RDF into the web.
- Publishing such descriptions on the web creates the semantic web.
- RDF bridges with other markup languages and transform them. It is possible to link RDF via meta header element in XML.
- The semantics of RDFS, OWL Lite, etc. are based on knowledge representation algorithms.
Ontologies/rules:
- They have to be created to frame the data of the area to be described.
- Support decentralized definition and management, across the network. It relies on a democratic bottom-up control of terms.
- Librarians have some expertise in making ontologies.
With SPARQL engine more powerful queries can be asked.
It’s interesting to know that:
- Smarter machines teach the computers to infer the meaning of web data.
- Smarter data makes data easier to find, access and process for machines.
- With machines having done this we have: extra rigor (naming relationships), extra information (identyfing relationships). No attached user-interface action.
For more information: http://www.w3.org/People/Ivan/CorePresentations/SW_QA/Slides.pdf
1 Globally unique identifiers for retrieval and universal naming tools.