JiscMail Logo
Email discussion lists for the UK Education and Research communities

Help for DC-GLOSSARY Archives


DC-GLOSSARY Archives

DC-GLOSSARY Archives


DC-GLOSSARY@JISCMAIL.AC.UK


View:

Message:

[

First

|

Previous

|

Next

|

Last

]

By Topic:

[

First

|

Previous

|

Next

|

Last

]

By Author:

[

First

|

Previous

|

Next

|

Last

]

Font:

Proportional Font

LISTSERV Archives

LISTSERV Archives

DC-GLOSSARY Home

DC-GLOSSARY Home

DC-GLOSSARY  May 2011

DC-GLOSSARY May 2011

Options

Subscribe or Unsubscribe

Subscribe or Unsubscribe

Log In

Log In

Get Password

Get Password

Subject:

Re: New User Documentation pages are on the wiki!

From:

Pete Johnston <[log in to unmask]>

Reply-To:

Pete Johnston <[log in to unmask]>

Date:

Tue, 3 May 2011 18:03:48 +0000

Content-Type:

text/plain

Parts/Attachments:

Parts/Attachments

text/plain (53 lines)

Hi Tom, all,

> 2) http://wiki.dublincore.org/index.php/User_Guide
>    The User Guide, ported from the MPDL Colab wiki!

In the section "What is Linked Data?"

"Metadata are the backbone of this method, making statements about data and how they relate to each other." 

I think we're really dealing with how things/resources (not just data) relate to each other. I wonder whether really it may be better to try to sidestep the whole metadata/data issue at this point? So maybe just say something like

The backbone of this method is making statements about resources and how they relate to each other

"Based on the above said a metadata description in Linked Data consists of: "

Particularly in the bullet list following here, I worry slightly we're mixing concepts from RDF and concepts from the DCAM ("description", "value", "value string" etc), and it may be a bit confusing for someone having read other introductions to "Linked Data". 

I'm not sure I have an easy answer to this, other than by expanding things a little bit to couch everything in terms of the RDF triple model, and saying explicitly that e.g. a "description" is a set of triples, talking more explicitly about the rdf:type property, the dcam:memberOf property and so on. This is sort of implicit in the presentation of the graph following this bullet list, but I think it might help to make it more explicit.

And I guess this is just a particular example of this issue, but I think there is some inconsistency in the use of the term "value". At the start

the object - the part that identifies the value this property has when describing this thing

so here the "value" is something identified by the object.

In the longer bullet list

values: bits of descriptive information (string literals) or references to other entities (resources), such as people or concepts, which are related to the described resource via the properties.

Here a "value" is a part of the graph/description, a literal or a reference to an entity/resource.

references to vocabulary encoding schemes (VES), enumerated sets of resources of which the things referenced as values are members

Here, (I think?) we're back to "value" being the thing not the reference?

In

http://wiki.dublincore.org/index.php/User_Guide/Creating_Metadata

having introduced the triple model in the introduction page, should this page be more explicit about how the tabular convention corresponds to the triple model?

I must admit I find the "red string" convention a bit confusing. Clicking through to the Turtle, I can see it's for the object="blank node" case, but I don't think it's used consistently e.g. compare the two Turtle examples, both with blank nodes

http://wiki.dublincore.org/index.php/User_Guide/Publishing_Metadata#exTit1
http://wiki.dublincore.org/index.php/User_Guide/Publishing_Metadata#exTit2

but in the tabular examples use different conventions (white rows and red string). I think the difference is because of the repetition, but in the triple model these are both handled exactly the same way, if you see what I mean.

Under "Guidelines for the creation of content for agents", the intent is to show the use of a URI to identify the agent, i.e. in the Turtle gnd:39454-3 is a QName which will get mapped via a prefix mapping to a URI, but from the tabular view it is quite hard to work out that that is the intent. 

I guess these are just symptoms of the general point that I think we need to define clearly how that tabular convention represents the triples, and then follow that convention carefully.

Pete

Top of Message | Previous Page | Permalink

JISCMail Tools


RSS Feeds and Sharing


Advanced Options


Archives

September 2011
June 2011
May 2011
March 2011
December 2010
November 2010
October 2010
September 2010
August 2010
June 2010
May 2010
April 2010
March 2010


WWW.JISCMAIL.AC.UK

Secured by F-Secure Anti-Virus CataList Email List Search Powered by the LISTSERV Email List Manager