Semantic Agent Systems

Lecture Notes in Computer Science, 2004

We report on experience with using RDF to provide a rich content language for use with FIPA agent toolkits, and on RDFS as a metadata language. We emphasise their utility for programmers working in agent applications and their value in Agent-Oriented Software Engineering. Agent applications covered include Intelligent Information Agents, and agents forming Virtual Organisations. We believe our experience vindicates more direct use of RDF, including use of RDF triples, in programming knowledge architectures for a variety of applications.

IGI Global eBooks, 2011

As Web-based content becomes an increasingly important knowledge management resource, Webbased technologies are developing to help harness that resource in a more effective way. The current state of these Web-based technology-the "first generation" or "syntactic" Web-gives rise to well-known, serious problems when trying to accomplish in a non-trivial way essential management tasks like indexing, searching, extracting, maintaining, and generating information. These tasks would, in fact, require some sort of "deep understanding" of the information dealt with: In a "syntactic" Web context, on the contrary, computers are only used as tools for posting and rendering information by brute force. Faced with this situation, Berners-Lee first proposed a sort of "Semantic Web" where the access to information is based mainly on the processing of the semantic properties of this information: "… the Semantic Web is an extension of the current Web in which information is given well-defined meaning [emphasis added], better enabling computers and people to work in co-operation" (Berners-Lee, Hendler, & Lassila, 2001, p. 35). The Semantic Web's challenge consists then in being able to manage information on the Web by "understanding" its proper semantic content (its meaning), and not simply by matching some keywords.

Encyclopedia of Complexity and Systems Science, 2009

Dereferencing a URI is using the URI to identify the object or resource it refers to. Folksonomy 'Folksonomy' is a neologism applied to structures that emerge from the practice of 'tagging' Web content. In some Web 2.0 applications, users can apply a tag (a descriptive term) to content such as a photograph or video clip. The tags need only be meaningful to the individual tagger, but if a large enough number of users tag content, descriptive structures analogous to more formal ontologies can emerge that are meaningful to wide communities. GRDDL Gleaning Resource Descriptions from Dialects of Languages (GRIDDL, pronounced 'griddle') is a mechanism for helping bootstrap the Semantic Web, by extracting RDF from XML documents, using transformations expressed in XSLT. GRDDL became a W3C recommendation in 2007. Metadata Metadata is data about data. In the context of the Semantic Web, metadata are also called 'markup' or 'annotations'. Because one aim of the Semantic Web is to support machine processing of information, metadata are helpful in describing the content of data. For example, metadata attached to a series of numerals could explain that it represents a zip code or a height or a population figure. Ontology An ontology defines, describes and constrains the concepts and relationships that are used in some particular domain of knowledge. Ontologies may therefore have an important role in data sharing, for example by providing a means of expressing which concepts (in the ontology) are referred to by particular terms (in a set of databases, which may use widely differing vocabulary). OWL The Web Ontology Language (OWL) is a language for describing and sharing ontologies on the Web. It is an extension of RDF, and has three variants. OWL Full is maximally expressive and compatible with RDF (any legal RDF document is a legal OWL Full document), but is undecidable. OWL DL is based on a series of restrictions of OWL Full which support efficient reasoning, at the cost of losing the strong connection with RDF (not all RDF documents are legal OWL DL documents). OWL Lite is even more restricted in expressivity, but is easier to grasp. OWL became a W3C recommendation in 2004. RDF The Resource Description Framework (RDF) is a standard framework for representing data on the Web, representing it in a three-place relation, of subjectrelation-object form, called a triple. It uses URIs to refer not only to the two items related, but also to the relationship asserted between them. In this way it extends the linking structure of the Web by allowing the nature of the link asserted to be described. Its syntax is XML-based, to support syntactic interoperability between the two. RDF became a W3C recommendation in 1999. RDF(S) RDF Schema (RDF(S) or RDFS) is a language for representing information on the Web. It is an extension of RDF to allow the description of the relationships which can be asserted between resources using RDF. So, for instance, RDF allows the assertion of the relationship 'author' between, say, 'Herman_Melville' and 'Moby_Dick', but RDF(S) is required to assert the properties of the 'author' relationship (e.g. that every document has at least one author, or that the 'author' relationship is the inverse of the 'written_by' relationship). RDF(S) became a W3C recommendation in 2004. Rules and RIF Rules govern the transformation of, and inference from, data. In particular, when data are being shared, it may be useful to make basic inferences over the data (for example, to determine whether two names refer to the same object or different objects). Rule-based knowledge such as this cannot be expressed in individual data stores, and may be hard to express in an ontology. The Rule Interchange Format (RIF) is a language, not complete at the time of writing, to express the most common or basic types of rule. SPARQL SPARQL is a special query language designed specifically for querying data stored in RDF ('SPARQL' is a recursive acronym standing for 'SPARQL Protocol And RDF Query Language', and is pronounced 'sparkle'). At the time of writing (December 2007), SPARQL was a candidate recommendation by the W3C. Triples and Triplestores A triple is a statement in RDF, consisting of a subject, an object, and a binary predicate that relates them. Each item of the triple is identified by a URI reference (or can be a string literal, such as a date or a number or name). A knowledge repository which contains RDF triples is called a triplestore. Triplestores may need to contain several millions of triples, and so need to be able to support fast querying at these potentially very large scales. URI A Uniform Resource Identifier (URI) is a string of characters for identifying an abstract or physical object or resource. There are different URI schemes (i.e. different ways to restrict the syntax of a URI to make it meaningful). A Uniform Resource Locator (URL) is a particular type of URI that identifies its object by means of its access mechanism or 'location' in the network. URIs are important in that they act as a standard way of referring to objects on the Web.

2002

We describe the technical choices and the design of a multi-agents software architecture to manage a corporate memory in the form of a corporate semantic web. We then present our approach to tackle a distributed memory and distributed queries.

IEEE Internet Computing, 2000