Mr William Francis1, Mr Alex Ip1, Dr David Lescinsky1, Ms Margie Smith1, Mr Andrew Turner1
1Geoscience Australia, Symonston, Australia
Persistent identifiers (PIDs) at the feature level are critical foundation for data that is referenced in open distributed systems. System interoperability depends on such identifiers to be stable and provide machine-readable views in reliable profiles and formats. PIDs and their views are published using the pyLDAPI utilities (an opensource LinkedData API available at https://github.com/RDFLib/pyLDAPI). GA is now implementing the new version of pyLDAPI, including the innovation of content negotiation by profile – separate from media type, standardising these profiles and the expression of relationships to other features.
Profile templates are implemented in entity specific pyLDAPI implementations with data transformed and delivered as standard formats and profiles, sourced from an internal, organisational point of truth – a common organisational data model with well-defined internal identifiers.
Governance of persistent identifiers and their profiles is managed through organisational policy with entities and relationships conceptualised in ontologies and business rules are captured and implemented in tools such as the Persistent Identifier Service.
Applications using these mechanisms to publish such open data at the feature level granularity and their relationships include the geoscience AusGIN portal (http://www.geoscience.gov.au/) and Location Index initiative (LOC-I).
William Francis is the Data Architect at Geoscience Australia.
This poster will present on the range of approaches at Geoscience Australia in the informatics domain to improve the value and usability of open data by providing feature level persistence and standardisation of relationships.