[ckan-discuss] Next version of the LOD cloud diagram. Please provide input, so that your dataset is included.

William Waites william.waites at okfn.org
Wed Sep 8 23:18:12 BST 2010


On 10-09-08 22:33, John Bywater wrote:

> Perhaps mistakenly, I would hazard a guess that it's just a form field
> issue.

No, it is a data model issue.

> Could I ask, which form field (or fields) would be preferred instead of
> the existing email form field?
> 
> Would a free text field be preferable? Or is there an RDF form which
> allows exactly the balance of degrees of constraints and degrees of
> freedom which would satisfy each and every user, regardless of what kind
> of thing they are trying to record?

Yes there is. First you decide on a predicate (attribute name). You have
to decide on what you mean by author. A natural choice might be
"creator" from the Dublin Core terms [0].

The range of dc:creator is is dc:Agent. So your value is a URI that
identifies the Agent. This behaves somewhat like a foreign key into
another table. The agent has properties that describe it such as its
email address, home page, name, etc., whatever is required.

You don't constrain which properties the Agent might have a priori
in the data model - it is always possible to add another predicate,
indeed someone somewhere else can add another predicate without
consulting you. This is an important point. The data is *not* kept
only in your storage. You might constrain the list for convenience
in making an HTML form to enter information about the agent.

Some predicates enable you to disambiguate Agents that might have
been given different URIs for some reason, they are called inverse
functional properties. foaf:mbox is one of them, foaf:openid is
another. But this is a more complicated topic, let's get the basics
down first.

[0] http://dublincore.org/documents/dcmi-terms/#terms-creator


-- 
William Waites           <william.waites at okfn.org>
Mob: +44 789 798 9965    Open Knowledge Foundation
Fax: +44 131 464 4948                Edinburgh, UK

RDF Indexing, Clustering and Inferencing in Python
		http://ordf.org/



More information about the ckan-discuss mailing list