Intelligent Systems: HOME
Semantic Web

As the amount of information available online and in company information systems continues to grow exponentially, organizing this information and finding it, and leveraging this information to extract real value, becomes ever more challenging. Integrating information and systems between companies, departments, and systems is a nightmare. Meanwhile, IT costs continue to mount and fielding the new features and innovations to bring value to your customer seems to take ever longer.

The key to organizing, finding, and integrating information between diverse organizations and systems, and getting maximum value from this information is to represent information in a standard yet flexible representation, in other words to speak the same language. Luckily, the W3C has defined a standard for representing such information. The key is a language called RDF (Resource Description Framework) which represents all information as simple statements which take the form of triples (subject, predicate, object). This provides a simple, yet powerful, language for expressing knowledge about objects and relationships between objects in a standard way. In addition, they have defined standards such as OWL and RDFS on top of RDF to define standard vocabularies for common needs such as defining child-of relationships, property definitions, and the like. Taken together, these standards help realize a far reaching vision known as the Semantic Web. This vision, originally articulated by Tim Berners-Lee, usually considered the father of the World Wide Web, applies many of the ideas that were responsible for creation and explosive growth of the web itself (standard language for documents, addresses, and protocols, interconnections between documents, distributed self publishing on a massive scale) to the domain of knowledge representation, with the hope of achieving the same explosive growth and inter-connectedness of knowledge.

In addition to representing enterprise knowledge and metadata, and supporting system integration and inter-operability, RDF is starting to play a key role in making search more effective on the internet. Google and others are relying increasingly on the use of semantic markup in web pages to provide better search results based upon the meaning of content rather than just its raw text and surface syntax, as well as presenting search results in a structured format that is better suited to the needs of the searcher.

Intelligent Systems has been making the ideas and technologies of the Semantic Web, and in particular RDF, a central part of its technical architectures and solutions for many years. In addition, we have developed many tools for collecting, storing, and using knowledge and metadata represented as RDF in RDF-based knowledge and metadata repositories.

Some solutions we have developed and tools we have created to support these solutions over the years include:

  • Using RDF as an intermediate representation and lingua franca for content and metadata in content migrations and Content Management projects, as well as the migration tasks and workflows making up these migrations.
  • Using RDF to represent acquired knowledge and metadata in Crowdsourcing tools, as well as the knowledge required to drive the Crowdsourcing tool itself (e.g. questions, answers, users, scoring, workflow, etc)
  • Using RDF to represent users, products, and related metadata in personalization and recommendation systems
  • Using RDF to represent knowledge in conjunction with rule-based inference engines to infer new facts, transform content, and generate targeted content in applications such as rule-based content migrations, recommendation engines, personalization systems, generating targeted content, to name a few.
  • Using RDF to represent tasks, workflows, data transformation pipelines, configuration information in many of our tools including the knowledge/metadata repositories themselves (Metador, Crowd Sorcerer) and content migration tools
  • Parsing and representation of Sun's Java API (Javadoc) in RDF representation to support semantic searching and relationship navigation in API documentation for Sun
  • Development of the Metador Knowledge and Metadata Repository tool
  • Development of the Crowd Sorcerer tool which is an RDF repository and crowdsourcing framework which uses RDF for iterative semantic crowdsourcing