Governmental Linked Open Data: A Data Management Perspective
1. Governmental Linked Open Data:
A Data Management Perspective
(or: what is in a link?)
Profa. Maria Luiza Machado Campos
Federal University of Rio de Janeiro
mluiza@ufrj.br
5. Open Governmental Data
• In some sense
– There is a lot already available
– But…
• Is it easy to find, access and use together?
6. Open Governmental Data
• Interoperability and Integration requires a lot of
work
– Even inside a single organization!
(We’ve had DBMS for more than half a century, but have we
achieved the integration we aimed at? What was missing?)
– And now, on the Web:
• Distributed, heterogeneous, in large scale, highly visible,
large number of different users, …
7. Linked Open Data (LOD)
– What is it all about?
• Using standards
• A very “fine-grain” representation (RDF triples)
– Enables LINKING with flexibility!
– Simple?
– Powerful!
• It has interoperability in its essence
– RDF: Resource DESCRIPTION Framework
» Created to interoperate METADATA!
– Common understanding
8. Linked Open Data (LOD)
– An exciting “new” way of publishing and
consuming data!
• The power of linking
• The power of collaboration
– Consumers are publishers too!
• More and more data being generated and linked
– Sensors
– Web of things
• Data and Metadata being explored together
– Are some links more important than others?
– Links between types; Data hubs
• Querying AND Navegating AND Searching
9. Linked Open Governmental Data
– It is not just about converting to a new flexible
representation
– It is important to be “linkable”
– Descriptions are important
• To know what it really means
• To know where it comes from
• To know what we have available
• ….
– Some data are more difficult to describe than
others
10. Linked Open Governmental Data
– Open is good!
– Flexible is good!
– Semantically interoperable is even better!
• Having a “ data management approach” to Gov. LOD
– The role of vocabularies, glossaries
– The role of database schemas
– Creating/using existing complementary metadata
» e.g. Provenance metadata, voID
– Creating/Exploring ontologies as reference models of a domain
» Using Foundational Ontologies as a common ground
A data quality – oriented Governmental LOD
11. But how long will it take?
Will it ever happen?
– Lessons learned from the past
• Incremental, evolutionary
• Agile, but planned
– Estimulate initiatives
• Learn by practicing!
– Pilot projects
• Evaluate feedback and impact
– Observe standards and best practices from existing
initiatives
12. Linked Open Governmental Data
– Plan and act for the next steps
• Capacitate
– Not only on available technologies, but on reflecting about their use
– There is a lot to be learned from different knowledge areas
» Semantic web + databases and data management + AI +
• What has worked, and also what has not worked
• Define standards and strategies
– In Brazil:
» E-Ping, INDE, INDA
13. Linked Open Governmental Data
But can we what have we learned from data management
in the past?
• To define priorities
– There are very different kinds of data
– Master and reference data need special attention
– Statistical data is not trivial
» Faceted or multidimensional, sampled, spatio-temporal
heterogeous
• To define strategies for building good reference
models/ontologies
– They are fundamental for interoperability
– Conceptual modelling is important!
• To support maintenance, lineage, evolution
– LOD in the data management process of govern. institutions
14. Where universities can help
• Forming
• Developing, prototyping and experimenting
new technologies, approaches, innovative
applications
• Partnerships (with government and private
initiative)
– Applying new approaches and technologies
– Insights from different knowledge areas
– Discussing perspectives and practices
• Defining new and rescuing old trends!