ISBD for the Semantic Web: namespaces

Report
ISBD for the Semantic Web:
namespaces, elements, vocabularies, application
profile
Gordon Dunsire
Presented at Centar zu Stalno Stručno
Usavršavanje (CSSU), Zagreb
21 Nov 2011
Semantic Web
“machine-readable metadata”
Faster! 24/7/365! Global!
Metadata expressed as “atomic” statements
A simple, single, irreducible statement
The title of this book is “Treasure island”
In a standard machine-processable format
Resource Description Framework (RDF)
Resource Description Framework
Metadata statement constructed in 3 parts
“Triple”
The title of this book is “Treasure island”
Subject of the statement = Subject: This book
Nature of the statement = Predicate: has title
Value of the statement = Object: “Treasure island”
This book – has title – “Treasure island”
subject – predicate - object
Identifiers
Need unambiguous way of identifying each
part of the triple for efficient machineprocessing
Human labels (“This book”, “has title”) no good
Same thing, different labels; different things, same
label
Exploit the utility of the URL
Machine-readable, regular syntax, unambiguous
Uniform Resource Identifier (URI)
Uniform Resource Identifier
Can be any unique combination of numbers and
letters
No intrinsic meaning; it’s just an identifying label
Can look like a URL
http://iflastandards.info/ns/isbd/elements/P1001
But does not lead to a Web page (in principle ...)
RDF requires the subject and predicate of triple
to be URIs
Object can be a URI, or a literal string (“Treasure
island”)
Namespaces
URI can be constructed from a base plus a
unique, identifying suffix
http://iflastandards.info/ns/isbd/elements/
+ P1001
Base is known as a namespace
Can be abbreviated by human programmer
“isbd” = http://iflastandards.info/ns/isbd/elements/
isbd:P1001
Machine expands abbreviation for processing
Everything as triples in RDF
Every aspect of the metadata must be
expressed in RDF to be machine-processable
Metadata about real-world objects (books,
people, etc.)
Metadata about the predicates (definition, label,
scope, etc.)
Common predicates apply to many types of thing
(human-readable label, etc.)
High-level RDF namespaces (rdfs, owl)
RDF is expressed in RDF (“bootstrap”)
Creating namespaces and URIs
ISBD is using the Open Metadata Registry
Can assign a running “number” to the base to
create a new URI
Set of properties (= predicates) for creating
basic triples about other properties
E.g. rdfs:label for assigning a human-readable
label to an RDF property (or class)
isbd:P1001 - rdfs:label - “has content form”
A property of a property
Subject
Predicate
isbd:P1001
rdfs:label
Object
“has content form”
Subject
isbdcf:T1008
Predicate
skos:prefLabel
Object
“spoken word”
Application profile
Need a way to specify how a useful “record”
can be constructed from RDF triples
Which triples are involved, and from which
namespaces?
Sequence? Repeatable? Mandatory?
Sub-component aggregations
Publication statement = place + name + date
Content rules?
Mandatory
Not repeatable
Aggregation of simpler elements
Syntax of aggregation (punctuation)
Thank you
[email protected]
Open Metadata Registry
http://metadataregistry.org/

similar documents