También disponible en Español

Inf@Vis!

The digital magazine of InfoVis.net

Semantic Web
by Juan C. Dürsteler [message nº 26]

The Semantic Web is presented as the new Internet revolution. The promise is to convert information to knowledge.

The World Wide Web is an information space that has allowed new levels of human communication. For this very reason the information that it contains has been designed basically for human consumption and uses a language that makes it difficult for machines to exchange and effectively elaborate data.

E-commerce applications, for instance, require the flow of data between suppliers, distributors, retailers and even the end user. Currently the interchange consists of simple data transactions, many times just data separated by tabs or driven by very specific applications.

The vision that lies behind the Semantic Web is that the data that is to be in the Net be defined in such a way that machines could use it and even understand its meaning, without human intervention. But the thing goes beyond this. What is intended is to convert information into knowledge by coding the data with metadata, data about data, readable in automatic way.

This coding comes from the definition of different Ontologies. An Ontology is, in this context (not to be confused with the philosophical concept) the specification of a conceptualization, i.e. a set of definitions of concepts. The Ontologies are expressed by means of representation languages, like the Resource Description Framework (RDF) that, in the Semantic Web, is built on top of XML.

The creation of Ontologies is promoting the development of Metadata Editors or Ontology Editors like Protégé or Webonto   and systems to favour the interoperability between different Ontologies. There is also active work in the processing of them by means of Inference Engines that allow you to infer new knowledge about already specified knowledge.

The potential of this future Web is huge. In principle you could create software by just finding the proper components in the Net along with having the specification of how to link them. An appropriate agent (not necessarily a human one) could perform this operation.

But for the organisations it could be a real quantum leap by allowing them to code their internal knowledge and to use it properly for their relations with suppliers and customers through the Net.

The materialisation of this vision is supported, among others by Tim Berners-Lee (it's worth reading his papers on the topic) and the organisation SemanticWeb. To become a reality it will need standardisation, the incorporation of the many tools and technologies currently in development and the adoption of both by the market.

The first example of official standard is that of ISO/IEC 13250 on Topic Maps and there are already some companies that offer tools and consultancy, like Ontopia. The US electric power industry has adopted RDF to exchange power system models . Nevertheless there is still a long way to go.

On the other hand, the European Community is promoting research on this topic  within the V Research Framework.

The difference between knowledge and information is what makes an organisation effective. The Semantic Web could result in a giant leap in order to convert one into the other. For the moment we must wait and see

Links of this issue:

http://www.w3.org/RDF  
http://smi-web.stanford.edu/projects/protege/  
http://kmi.open.ac.uk/projects/webonto  
http://www.w3.org/DesignIssues  
http://www.semanticweb.org/   Semantic Web Org
http://www.topicmaps.org/  
http://www.ontopia.net/  
http://www.cordis.lu/ist/ka3/iaf/swt_workshop.htm  
© Copyright InfoVis.net 2000-2018