Content Management with RDF
Leigh Dodds, Chief Technology Officer, Ingenta
25th April 2007
Why?How?
A Digression: The RDF Model
<http://www.example.org/article/1>
Alan Turing worked at Bletchley Park
</person/turing> workedAt </place/bletchley>
</person/turing> hasName “Alan Turing”
</person/turing> workedAt </place/bletchley>
</place/bletchey> hasName “Bletchley Park”
< /p er s o n / tu r in g > "Alan T u r in g "h as N am e
< /p ers o n /tu rin g > " A lan Tu rin g "h as N ame
< /p lace/b letch ley >" B letch ley P ark " h as N ame
w o rk ed A t
< /p ers o n /tu rin g > " A lan Tu rin g "h as N am e
< /p lace/b letch ley >" B letch ley P ark " h as N am e
w o rk ed A t
< /s d s d s d s >
< /p d fd s fs d >
< /erey ty y rty >
< /p lab led fd ftch ly >
< /p lace/b letch ley >
" ab c"
" s d s d aa"
" fg ee " fg ee " fg ee
" s d d s b c"
" 1 0 "
" 1 0 "
< /p lab led fd ftch ly > < /p lab led fd ftch ly >
" s d d d d c"
" s 2 3 "
" 3 .2 "
Why Use RDF?• Encourages modelling of resources
• Simple extensibility
• Ease of Storage
• Data Merging
• Networked data
• Re-wiring publications
Where Would You Use RDF?
F u ll T e xtCo n te n t M etad ata
S ear c h Br o w s e
T em p la tin g
..and Where Wouldn’t You Use RDF?
It’s Not For Content!
How?
Assigning Identifiers
http://metastore.ingenta.com/article/456
Vocabularies
Dublin Core• Basic set of generic
properties
• Title• Creator• Publisher• Created, Updated, Modified• Format• Identifier• Subject
• ..etc
PRISM• Publishing Requirements for
Industry Standard Metadata • Builds on XML, RDF, Dublin
Core
• DOI, eISSN• startingPage, endingPage• isPartOf, hasParts• related• references• hasTranslation,• hasCorrection
FOAF• “Friend of a Friend”• Community driven
specification• Some useful stable terms
for describing people and organizations
• Person• Organization• Document• homepage, weblog• topic, primaryTopic
SKOS• Simple Knowledge
Organization System• Thesauri, subject
classifications• Drive navigation and
indexing
• Terms• broaderTerm,
narrowerTerm• relatedTerms• preferred label, alternate
label
Generating RDF• XML to RDF
– XSLT– XQuery
• Database to RDF– D2R
• XHTML to RDF– GRDDL
Manipulating RDF
< /p ers o n /tu rin g > " A la n T u rin g "h a s N a m e
< /p la ce/b letch ley>" B letch ley Pa rk" h a s N a m e
w orkedAt
< /s d s d s d s >
< /p d fd s fs d >
< /ereytyyrty>
< /p lace/b letch ley >
< /p lace/b letch ley >
" a b c"
" d ef"
" fg ee " fg ee " fg ee
" s d d s b c"
" 1 0 "
< /p ers o n /tu rin g > " A la n T u rin g "h a s N a m e
< /p la ce/b letch ley>" B letch ley Pa rk" h a s N a m e
w orkedAt
< /s d s d s d s >
< /p d fd s fs d >
< /ereytyyrty>
< /p lace/b letch ley >
< /p lace/b letch ley >
" a b c"
" d ef"
" fg ee " fg ee " fg ee
" s d d s b c"
" 1 0 "
SPARQL• W3C Query Language for RDF
SELECT ?resource, ?name
WHERE {
?resource hasName ?name
}
?resource ?name</person/turing> “Alan Turing”
</place/bletchely> “Bletchley Park”
Problem Areas
Duck Typing for Data
Questions?
Picture Attributions“The Confusion of Tongues”, Gustav Doré
– http://en.wikipedia.org/wiki/Image:Confusion_of_Tongues.png
“So many ducks…Ducking Hell”, Gaetan Lee– http://flickr.com/photos/gaetanlee/298160427/
