danbri wrote (22/03/2009)
Dan, this is certainly on our current agenda. We thought it would be easier to have a SKOS discussion if we were able to publish a 1000 number outline as a real demonstrator on the web. We could have published the simple text selection of the outline in October - but when I looked into it I realized that classes on this level (i.e. 1000 numbers out of 70,000) do not normally have all details necessary to demonstrate the number synthesis rules. For instance on the top 1000 subdivisions in the UDC we do not have examples of synthesis such as
33:32 Relationships between politics and economy
or more precisely
33-042.3:32 Influence of politics on economy
Phase relationships make sense in detailed indexing and not on the top level disciplines/knowledge areas. But in order to demonstrate this functionality we had to add such examples of combinations in the 1000 number outline.
In addition to this editing we also decided to add multilingual data (and mappings), and to allow editing of translations online. This added a bit of complexity to the interface etc.
We hope that in a couple of weeks we are going to have this up. We would then look into making these data available in various schemas, SKOS being one of them.
"How are things going with SKOSification? If some SKOS extensions (SKOS is designed to be independently extended) are needed to do justice to UDC, it would be great to begin that discussion... particularly as SKOS has just entered "candidate recommendation" stage at W3C - http://www.w3.org/News/2009#item35"
Dan, this is certainly on our current agenda. We thought it would be easier to have a SKOS discussion if we were able to publish a 1000 number outline as a real demonstrator on the web. We could have published the simple text selection of the outline in October - but when I looked into it I realized that classes on this level (i.e. 1000 numbers out of 70,000) do not normally have all details necessary to demonstrate the number synthesis rules. For instance on the top 1000 subdivisions in the UDC we do not have examples of synthesis such as
33:32 Relationships between politics and economy
or more precisely
33-042.3:32 Influence of politics on economy
Phase relationships make sense in detailed indexing and not on the top level disciplines/knowledge areas. But in order to demonstrate this functionality we had to add such examples of combinations in the 1000 number outline.
In addition to this editing we also decided to add multilingual data (and mappings), and to allow editing of translations online. This added a bit of complexity to the interface etc.
We hope that in a couple of weeks we are going to have this up. We would then look into making these data available in various schemas, SKOS being one of them.
Thanks for the detailed reply. This makes perfect sense...
ReplyDeleteI'm getting very interested in how these kinds of synthesis rules relate to what we have in RDF/OWL, but am far from having a coherent story. When these UDC sections are released, I think that will help focus discussion...
Guess you saw http://www.w3.org/News/2009#item140 and http://www.worldcat.org/devnet/wiki/DeweyInfoTechOverview ... I want to blog about the SKOS news, anything I can say re UDC?
ReplyDelete