The Semantic Web vision described by Berners-Lee, Hendler and Lassila (2001) represents an approach to enhancing the current World Wide Web with machine-understandable semantics. The essential idea is that Web resources identified by Uniform Resource Identifiers (URIs) can be described by metadata with the purpose of enabling automated processing. The precision of logics-based description would allow for the creation of tools that do not rely on natural language processing as current search tools, thus ideally removing the problems associated with ambiguity and implicitness in natural language. Metadata requires shared semantics in the context of an open Web, so that the proliferation of descriptions and vocabularies do not end up in a Tower of Babel. Ontologies provide the support for the shared representation of such semantics, and ontology mapping procedures could resolve the issues of semantic interoperability in case of disparate representations over the same domain. Ontologies act as descriptions of semantic domains used to express metadata records according to previously agreed definitions. Based on the infrastructure provided by ontologies and semantic metadata, the Semantic Web vision rests in the provision of distributed functionality built on intelligent agents and Web services. Since a virtual community is no other thing than a phenomenon that takes place on the Web, its elements are thus subject to being described by Semantic Web metadata, so that community-specific Semantic Web functionality could enhance the communicational elements that are part of communities. A typical scenario for such functionality is that of a filtering service that delivers recommended resources for community members according to the interests shared by members with similar interests. Other scenarios may result in the adaptation of the shared space inhabited by the community members; for example, member lists may be reordered according to the strength of social relationships, or messages could be circulated according to the network of social relationships (Sicilia & García, 2004). The Semantic Web opens many perspectives to create richer environments for virtual communities, and some specific technologies have already been created to deal with some of the definitional aspects of communities. Here we sketch the main elements that should be taken into account in developing experimental applications of Semantic Web technologies to the specifics of virtual communities.
Training Intelligent Agents in the Semantic Web Era: The Golf Advisor Agent