CEF-søknad:

STIRData: Specifications and Tools for Interoperable and Reusable data

Faglig arena for informasjonsforvaltning 10. mars 2020

​

Steinar Skagemo

seniorrådgiver

Direktørens stab

​

Problem: Data quality and Availability

• “Poor quality and availability of Open Data have been acknowledged as the most important technical barriers which hamper the reuse of open data”
• Relevant factors affecting quality:
• disparity of data formats and structures followed by different data providers (different models, terms list etc)
• inconsistencies in catalogue descriptions (metadata)
• incompleteness of the actual content (content about same entity spread across many datasets, not connected)
• Searches in European Data Portal characterised by low precision or low recall
• Significant efforts in building APIs for re-use of certain High Quality Datasets in each country does not solve re-use across borders

Builds on ...

• ISA², The Semantic Interoperability Community (SEMIC) and Access to Base Registries (ABR): Work on specifications in the form of Core Vocabularies
• BRIS, TOOP, euBusinessGraph, DCAT-AP
• Discussions on challenges and good practices identified in the framework of SEMIC
• Specifically: Linked Data Showcase Pilot (2019)
• Fra Norge: Difi, Brønnøysundregistrene
• Presentert på SEMIC 2019: SEMIC 2019: Panel on the Linked Data Showcase (LDS) pilot [PPT]
• Case: Business Registries / Company Data

Linked Data Showcase Pilot

Ansvaret for ulike autoritative data er distribuert -- både innenfor og på tvers av organisasjoner og landegrenser

Kartverket

Brønnøysundregistrene

Men bruken av dataene går ofte på tvers av ansvarslinjene - det kreves data som forvaltes av mer enn en ansvarlig

Regnskaps-registeret

Enhets-�registeret

Administrative grenser

The Business Register in Belgium

Business Registry

Uten å gjøre en manuell integrasjon på tvers av ulike kilder, ønsker jeg mer informasjon om en enhet, som f.eks. regnskap, kommune, region, internasjonalt hovedkvarter etc

Annual Accounts

Tim Berners-Lees prinsipper for lenkede data - og de tre reglene

1. Alt vi kan snakke om (fysiske og abstrakte “ting”) har navn som starter med HTTP
2. Du kan bruke HTTP-navnet til å slå opp og få data i retur
3. Dataene du får i retur uttrykker relasjoner gjennom nye HTTP-navn

STIRdata proposes:

• “… the use of Linked Data and semantic technologies …”
• “ … to ease the re-use of … data from multiple sources”
• “ … deliver more holistic and analytical data insights and pave the way for automation and added-value services … “
• “ … without changing the distributed responsibility for data-governance.”
• “To ensure high quality data, it is of high importance that the data "lives" in the same organisation that has the highest interest in keeping the data up to date and relevant, to fulfill their mission.”

Oppsummert: Løse utfordringen med datakvalitet og tilgjengelighet ved å bruke teknologi som støtter “forvalte lokalt, bruke globalt”

Contribution to Norwegian National Strategy

• “One Digital Public Sector” launched in June 2019 states as one its six goals:
• “The public sector shall exploit the potential of sharing and using data to create user-friendly services, and to promote value creation in the business sector”
• Specific actions:
• “work on “a common methodology, set of principles and framework for a generic «data distributor»”

Outcome

1. User Scenarios
2. Set of specifications and guidelines
3. A deployable framework of tools [build on LinkedPipes ETL, D2RML Processor]
4. Improve the content available via the European Data Portal and National Data portals
5. Demonstrate in practice the value and reuse potential through online platform for searching, navigating, visualising etc
6. A thorough evaluation of the proposed solutions and services

Consortium: Eight partners in four countries