Describing weblogs with RDF

Contents

1. Introduction
2. Namespaces
3. Classes
4. Properties

Introduction

The purpose of creating an RDF vocabulary for describing weblogs is to enable exchanges among systems which create or collect information about weblogs. If a system knows, for example, that Doc Searls includes David Weinberger in his “blogroll” (a list of weblogs he recommends to others), it can use RDF to express this fact in a way that other systems can understand.

<rdf:RDF
  xmlns:rdf="http://www.w3.org/1999/02/22-rdf-syntax-ns#"
  xmlns:blog="http://www.eyrie.org/~zednenem/2002/weblog-rdf/">
  
    <blog:Weblog rdf:about="http://doc.weblogs.com/">
        <blog:recommends rdf:resource="http://www.hyperorg.com/blogger/"/>
    </blog:Weblog>

</rdf:RDF>

The goal is to be able to express every weblog-specific fact about any given weblog using this vocabulary. This includes blogrolls, what posts are current, what weblog a post is part of, authorship, publication dates, and so forth. It builds upon the web threading vocabulary by creating a new set of names to describe weblog-specific information. Like the web threading vocabulary, it also recommends the Dublin Core vocabulary for describing cataloguing and publishing information.

Using RDF makes it trivial to provide additional information for weblogs, archives, and posts. For example, a syndication vocabulary could be applied to weblogs to indicate how frequently they are updated or a moderation vocabulary could allow people to rank posts by quality.

As an example, this description of ZedneWeb circa 2002-07-03 demonstrates many of the classes and properties described here.

Namespaces

This convention uses three vocabularies to describe weblogs. The prefixes given here are only for the purposes of this document. In real-world RDF, any value may be used for the prefix, so long as it is declared with the appropriate URI.

Prefix	URI	Comment
thread:	http://www.eyrie.org/~zednenem/2002/web-threads/	Describes threaded resources such as weblogs and message boards
blog:	http://www.eyrie.org/~zednenem/2002/weblog-rdf/	Additional information specific to weblogs
dc:	http://purl.org/dc/elements/1.1/	The Dublin Core; provides cataloging and publication information

For more information about the Dublin Core, see the reference description. The representation of Dublin Core elements is based on the suggested RDF encoding

Classes

These three classes describe the components of a weblog.

Class	Comment
blog:Weblog	The primary page of a weblog, used to represent the weblog as a whole.
thread:Archive	A page containing one or more posts in their permanent locations.
thread:Post	An individual weblog post.

Properties

The properties below are used to describe weblogs and their component parts. When given, the “Domain” column lists classes which the property can describe and the “Range” column lists classes which can be the value of the property.

Property	Comment	Domain	Range
dc:creator	A string describing the authors	blog:Weblog thread:Post [1]
dc:date	A string giving the publication date [2]	thread:Post [1]
dc:title	The title of a post or weblog	blog:Weblog thread:Post [1]
blog:currentPosts	An ordered lists of post considered to be current	blog:Weblog	rdf:Seq
blog:hasArchive	An archive of posts from the weblog	blog:Weblog	thread:Archive
blog:inWeblog	Indicates membership in a weblog	thread:Archive thread:Post	blog:Weblog
blog:recommends	An item to which the weblog has a prominent link	blog:Weblog
blog:linksPage	A page containing a list of featured resources	blog:Weblog
blog:rssChannel	The location of an RSS channel associated with the weblog	blog:Weblog
thread:agreedWithBy	A post which agrees with this resource		thread:Post
thread:agreesWith	A resource which the post agrees with	thread:Post
thread:commentedOnBy	A post which discusses this resource		thread:Post
thread:commentsOn	A resource which the post discusses	thread:Post
thread:disagreedWithBy	A post which disagrees with this resource		thread:Post
thread:disagreesWith	A resource which the post disagrees with	thread:Post
thread:followedUpBy	A post which corrects or provides new information	thread:Post	thread:Post
thread:followsUp	An earilier post which this post provides new information about	thread:Post	thread:Post
thread:hasPost	A post in the archive or weblog	blog:Weblog thread:Archive	thread:Post
thread:inArchive	The archive containing the post	thread:Post	thread:Archive
thread:next	The post or archive following this one	thread:Archive thread:Post	thread:Archive thread:Post
thread:pointedToBy	A post which refers to this resource but does not discuss it		thread:Post
thread:pointsTo	A resource which the post refers to but does not discuss	thread:Post
thread:prev	The post or archive preceeding this one	thread:Archive thread:Post	thread:Archive thread:Post
thread:quotedBy	A post which quotes this resource		thread:Post
thread:quotes	A resource which the post quotes	thread:Post
thread:refersTo	A resource the post refers to	thread:Post
thread:referredToBy	A post which refers to this resource		thread:Post

Note [1]: In general, dc:creator, dc:date, and dc:title can be applied to any resource. These listings are merely a suggestion of classes relating to weblogs where they make sense.

Note [2]: Timestamps will be formatted according to the W3C date/time note, which describes a profile of ISO 8601. Permissable granularities are day (2002-07-05), minute (2002-07-05T15:00Z), and second (2002-07-05T15:00:00Z). Times with reduced granularity are taken to mean “sometime during that period”. Thus, it cannot be said whether a post dated 2002-07-05T15:00Z comes before or after one dated 2002-07-05, only that they were published on the same date.