Web 3.0 for beginners

Mark Fisher — Mon, 22 Jun 2009 20:27:59 +0000

In this series of articles I will explain the basics of Web3.0, often called the semantic web, and how to start using future technologies now on your site.

I am assuming a basic knowledge of web and web2 technologies; these articles are intended to assist users in updating their projects to be ready for web3.

Basic concepts

URI

When using semantic information architectures, any artifact may be addressed by a symbol known as a URI or “unique resource indicator”. By artifact we mean any kind of information at all.

For example “rdf:type” is a property that may exist for any meta data; it is referenced by the following URI “http://www.w3.org/1999/02/22-rdf-syntax-ns#type”. In this case, the URI is also a URL meaning that you can go to this address in your browser (unless you can read xml/rdf or you are a machine, this will not make much sense). However, there is no requirement for a URI to address anything over the internet. nor for it to return anything meaningful if it does.

A URI is similar to a URL in that it has the same token structure. However a URL is assumed to be directly addressable over the internet, whereas a URI is an identifier. URIs will often start with “http:” – this is merely a convention and does not mean that any web content will exist at that location.

Meta-Data

Meta-data means information about information. Humans do not generally use meta-data themselves, however to classify documents using machines, we need an easy way to identify which information is contained in which document. We also need to be able to embed information in the document.

For example, say we have document X. Within X, part of the text will represent the title of the document, part will represent for example the author, and so on. A human may easily discern which part of the text refers to which type; however, for a machine there exists no easy way to distinguish parts of the document. For this reason we use meta-tags, either within the document or in a separate meta-document, to tell the machine that part A refers to the title, part B to the author, C to the date.

The semantic web exisits to define meta-data around data that we already have, and to create concept maps called ontologies, describing the various types of meta-data.

Ontologies

An ontology represents a classification system by which we may categorize meta-data. For example, the Dublin Core ontology contains properties such as “dc:title” and “dc:author” which refer directly to the title and the author of the document respectively.

Ontologies, and meta-data, are arranged in structures known as graphs. Within these structures we have two types of entity; nodes and arcs. Nodes represent individuals or items within the graph; arcs represent the relations between them. A node has an identifier (in RDF we use the URI to identify any unique instance); an arc has two nodes (in RDF we call these Subject and Object) and a predicate which specifies the type of relation,. For example, “rdf:instanceOf” represents a predicate which links an individual to the definition of the type that the individual represents.

A URI forms a functional identifier for a node; that is, any two elements with the same URI, refer to the same node.

Types

Any individual in Web3 must have a type; we specify the type using the “instanceOf” property (in rdf). Basic types are simple definitions, for example we may create a type “car” which has a supertype “automobile”. However the semantic web allows us to create more complex types using built-in rules; if our software has inferencing capabilities. we may make full use of these. Here is an example: we can define a type “4 wheeled objects” where the property “numberOfWheels” is equal to 4. If we create an individual, and assign it the property “numberOfWheels” with the value 4, our software will automatically infer that our individual is a member of the type “4 wheeled objects” with no further effort on our part.

Properties

Properties are assignments of some artifact, to another artifact. Two types of properties exists; object properties, where the value of the property represents another individual, and data properties, where the value of the property represents a known data type such as string or integer. Much like types, properties may be arranged in a heirarchy with the more general properties higher in the graph.

Individuals

An individual represents an instance of a type; this is similar to an “object” in oop. Basically, it represents a single “thing” rather than a type of thing.

Core Ontologies

There now follows a list of the most common ontologies

Dublin Core

Basic meta-data such as author, copyright, and other document related artifacts.

FOAF – Friend-of-a-friend

This ontology deals with relationships between people. For example, I could use “foaf:knows” in my document when linking to the homepages of people that I know.

SIOC Senantically Interlinked Online Communities

Similar to FOAF; allows me to specify metadata relating to social networking.

SKOS Simple Knowledge Organization System

Deals with concepts, and allows classification by grouping concepts into schemas. One may also specify semantic relationships between concepts; for exmaple, the “skos:broader” property links to a conecpt more general that the concept it links from (ie trees->plants, kilogram->wieght measurement, fly->insect).

Hopefully this basic guide will start you off; in the next part of the series, I will explain the various formats that may be used to express semantic meta data, and show you how to start using these right away in your web content.

Similar Blog & News Articles: Powered by

Similar Products: Powered by

	Semantic Web For Dummies (For Dummies (Computer/Tech)) :: Amazon Semantic Web technology is already changing how we interact with data on the Web. By connecting random information on the Internet in new wa
	A Semantic Web Primer (Cooperative Information Systems) :: Amazon The development of the Semantic Web, with machine-readable content, has the potential to revolutionize the World Wide Web and its use. A Sem
	Semantic Web Technologies: Trends and Research in Ontology-based Systems :: Amazon The Semantic Web combines the descriptive languages RDF (Resource Description Framework) and OWL (Web Ontology Language), with the data-cent
	Visualizing the Semantic Web: XML-based Internet and Information Visualization :: Amazon The Web has evolved from HTML quite dramatically over the last few years with revolutionary techniques for content and structural modeling,
	Programming the Semantic Web :: Amazon With this book, the promise of the Semantic Web -- in which machines can find, share, and combine data on the Web -- is not just a technical

Logicfish » Library and Information Science