Atom and the Semantic Web

Other pages:

Introduction

The regular HTML web is generally about delivering human-readable documents. The Semantic Web evolution is to make more machine-readable information available.

Atom is an XML-based system for weblogging and syndicated news delivery. Atom is based around metadata, and primarily uses HTTP, so it is reasonable to suggest that Atom is very much a Semantic Web application.

The expression of Atom in OWL is the most immediate way of linking Atom and Semantic Web technologies, but there are many issues involved. This document looks at the role Atom can play on the Semantic Web, and how Semantic Web technologies can help the development of Atom applications.

Status

This document has no official status. It ain't finished yet either.

Latest version : 2004-06-06

Comments

What is Atom?
What is the Semantic Web?
Common Misconceptions
What about RSS and the Semantic Web?
What is needed in Atom?
Resources
Glossary
Changes

1. What is Atom?

Atom is a system for distribution of information on the web. It is also a project that aims to create weblog and syndication formats and APIs that the developers can unite behind, allowing future progress in these fields without the political and technical problems that have plagued developers over recent years. The project was initiated by Sam Ruby, and is being developed collaboratively on a Wiki. There is also a public mailing list.

Rather than being text structured for layout, like HTML documents, Atom is structured for machine processing. A feed is comprised of a set of entries, and each entry corresponds to a resource. An (X)HTML representation of the resource (the content) will usually be provided, along with a set of metadata describing that resource. Essentially Atom is a resource description language. The Atom format is more content-oriented than the usual Resource Description Framework (RDF) - most of the time with RDF you're looking at URIs rather than inline XHTML-formatted documents (though this can easily be done).

Atom systems act as web services, but primarily using HTTP directly (i.e. REST) rather than the SOAP path, though full compatibility will be available.

2. What is the Semantic Web?

The Semantic Web is a vision of how the existing web can evolve into something considerably more useful.

The RDF model

Ontologies, OWL

SemWeb & Web Services

There is active development of Semantic Web Services, It is possible that the simple approach emerging from Atom could offer a baseline for lightweight delivery/access for Semantic Web Services.

Microcontent and the Lower-Case semantic web

Links

3. Common Misconceptions

Data needs to be in RDF format to be on the Semantic Web

For a start, it's misleading to talk about RDF format - RDF is a framework which at its heart has a relational data model, and this has several serialization syntaxes. The primary interchange syntax is RDF/XML, which looks complicated. But in virtually all cases, existing XML languages can be mapped directly to the RDF model and the data never needs to be expressed in RDF/XML. Another important point is that RDF/XML is primarily intended for machine-reading, other syntaxes (N3 and NTriples) exist that are considerably more human-readable.

The Semantic Web is a pipe dream

The Semantic Web is in the implementation phase. Many RDF-based systems are already online, and sophisticated systems using OWL are already appearing.