Transcript Presentation

Managing and publishing vocabularies
using a generic semantic registry
Simon J D Cox , Sally Tetreault-Campbell, Ben Leighton, Bruce Simons, Jonathan Yu
12 September 2016
LAND & WATER
Outline
•
•
•
•
Vocabulary formalization
Registration process
Linked Data Registry
BA Application
2 | Cox, Tetreault-Campbell, Leighton | Vocabulary registration
Terms and vocabularies,
linked data
3 | Cox, Tetreault-Campbell, Leighton | Vocabulary registration
4 | Cox, Tetreault-Campbell, Leighton | Vocabulary registration
SKOS: RDF for concept definitions
Simplified Knowledge Organization System
‘Term’ is a label for a concept,
not its identifier
Cox | Vocabularies & vocabulary services
For “vocabularies” with highly specific semantics
Geochronologic Era
Name = Neogene
Rank = Period
Geochronologic Era
Name = Pliocene
Rank = Epoch
Geochronologic Boundary
Name = Base of Messinian
Formally Ratified
Boundary Position
Geochronologic Boundary
Name = Base of Neogene
Cox | Vocabularies & vocabulary services
…define a highly specific concept!
7 | Cox, Tetreault-Campbell, Leighton | Vocabulary registration
Registration
8 | Cox, Tetreault-Campbell, Leighton | Vocabulary registration
Registration
Publish – make a resource available to target audience
Assign an identifier – or adopt one assigned elsewhere
Capture or record metadata – discovery/selection/use/status/provenance/…
following a set of rules and procedures
• Authorization
• Validation
• Scope
9 | Cox, Tetreault-Campbell, Leighton | Vocabulary registration
Registration terminology
• Repository
• A system to store resources
• Registry
• A system to issue and maintain
identifiers, and manage records
that describe resources
• Register
• A set of records managed on
behalf of a control body, in a
registry
• Catalogue
• Convenient interface to a set of
records, usually to assist discovery
10 | Cox, Tetreault-Campbell, Leighton | Vocabulary registration
Process and Roles
11 | Cox, Tetreault-Campbell, Leighton | Vocabulary registration
Source: ISO 19135:2005
Linked data registry
12 | Cox, Tetreault-Campbell, Leighton | Vocabulary registration
Developed by
Epimorphics for
WMO
13 | Cox, Tetreault-Campbell, Leighton | Vocabulary registration
and for DEFRA
14 | Cox, Tetreault-Campbell, Leighton | Vocabulary registration
now used for
Australian
Environment
Registry
Status and lifecycle – for registers & entities
16 | Cox, Tetreault-Campbell, Leighton | Vocabulary registration
Definitions may change over time
2004
2006
 Topology changes
Cox | Vocabularies & vocabulary services
2008
 Name changes
2010
LDR Versioning
Non-essential properties can be updated
E.g. definition text, relationships
Each change creates a new ‘version’ of the entity
Each version can be accessed forever
http://environment.data.gov.au/def/object/cadmium
http://environment.data.gov.au/def/object/cadmium:2
http://environment.data.gov.au/def/object/cadmium:1
18 | Cox, Tetreault-Campbell, Leighton | Vocabulary registration
Roles
19 | Cox, Tetreault-Campbell, Leighton | Vocabulary registration
Use for an evolving glossary
20 | Cox, Tetreault-Campbell, Leighton | Vocabulary registration
What are the impacts
of coal mining and
coal seam gas
extraction on water
resources?
http://www.bioregionalassessments.gov.au
21
Bioregional Assessments Programmme
•
•
•
•
Four agencies
200 people
13 bioregions, subregions
Five disciplines
− Ecology
− Hydrology
− Hydrogeology
− Geology
− Risk assessment
• Cross-cutting
− Information Management
− Products QA/QC
22
23
Large number of reports, tight editorial process
Bioregional assessments glossary
• Need common science terminology to harmonize across many
products – 200+ reports
• Definitions continually refined by science expert committee
• Each term in a report must be linked to the definition at the time
the report was written
BA Glossary
managed as a
register in the
Environment
Registry
Some
definitions
have had
multiple
revisions
27 | Cox, Tetreault-Campbell, Leighton | Vocabulary registration
Summary
28 | Cox, Tetreault-Campbell, Leighton | Vocabulary registration
What have I told you?
• Vocabularies can be formalized using RDF and published as linked data
• Registration is a process described with a formal model
• Linked Data Registry implements registration on a RDF/linked data platform
• May be used for managing any entity formalized in RDF
• Used to manage an evolving glossary for BA - additions and revisions
http://registry.it.csiro.au/def/environment
http://environment.data.gov.au/def
29 | Cox, Tetreault-Campbell, Leighton | Vocabulary registration
Thank you
Land and Water
Simon Cox
Research Scientist
t +61 3 9545 2365
e [email protected]
w www.csiro.au
LAND AND WATER
Definitions are
more than
SKOS
Definitions
linked to
well-known
systems where
possible
SKOS: RDF for concept definitions
The ‘term’ is a label for a concept,
not its identifier
Cox | Vocabularies & vocabulary services