Publishing Linked Data from Brazilian Politicians on the Web
Lucas de Ramos Araújo
Pablo N. Mendes
Jairo Francisco de Souza
At the Workshop on Semantics in Governance and Policy Modelling, Extended Semantic Web Conference 2011 ESWC2010.
May 30, 2011 - Crete, Greece
1. Publishing Linked Data from
Brazilian Politicians on the Web
Lucas de Ramos Araújo
Pablo N. Mendes
Jairo Francisco de Souza
At the Workshop on Semantics in Governance and Policy Modelling,
Extended Semantic Web Conference 2011 ESWC2010.
May 30, 2011 - Crete, Greece
3. Introduction: Motivation
● Why?
● Brazil has a substantial amount of government data
available on the Web, but usually not structured
●
Brazil ranks 61st in the United Nations E-Government
Survey 2010 losing 16 positions since 2008 [1].
● Brazilian citizens: over 90% of the interviewed have
average, little or no interest in politics [2].
● Mix information from different sources enable new
services and offer new insights about the people and
the government.
[1] United Nations e-Government Survey 2010 (http://unpan1.un.org/intradoc/groups/public/documents/un /unpan038851.pdf )
[2] Latin American Public Opinion (http://www.scielo.br/pdf/op/v15n2/09.pdf)
4. Introduction: Goal
● Our goal?
● Collect data from Brazilian politicians from different
sources and publish them as Linked Open Data.
● Enable reuse and interconnection between information
from different sources, thus generating new knowledge.
● Engage citizens in actively participating in the decisions
being taken by their elected government.
● Bridging a gap between the government and public
opinion.
5. Background
● Linked Open Data, incrementally:
★ make your stuff available on the Web under an open
license
★★ make it available as structured data
★★★ use non-proprietary formats
★★★★ use URIs to identify things, so that people can point
at your stuff
★★★★★ link your data to other data to provide context
Tim Berners-Lee's 5 star scheme. http://www.w3.org/DesignIssues/LinkedData.html
6. Data on the Web (1990s)
<
Approach: Integrate in-house, expose result of integration
BD
Integrated Portal Consumer
BD Data WEB
BD
Another Consumer
● One cannot anticipate all uses of his/her data
– For some, a portal is a nice way to access data
– For others, it is an obstacle on their way to create
interesting applications
7. Data on the Web (linked!)
<
Linked Data Approach: expose individual data, expose linkage
Portal
BD
Consumer
LINKED DATA WEB
BD
BD Another
Mashup
Integrated Another Consumer
Data
● A portal is just one option to look at the data.
– Third parties have the freedom to re-purpose the data
– Third-party perspectives may be useful for other people
– May be useful even for the publisher's portal!
8. Project Architecture
Data
Sources
Supreme
Electoral Court
RDF
Federal Senate
House of Screen Database LOD
Representatives
Scrapers Cloud
Brazilian
Politicians
HTML
Excellencies
15. Flexibility in Exploration
● Arbitrary queries on the data
● Representatives of states with high illiteracy that
vote on propositions related to 'education'
SELECT DISTINCT ?politico ?estado ?analfabetismo WHERE {
?politico a dbpedia-owl:Politician .
?politico dbpedia-owl:birthPlace ?estado .
?estado dbpprop:analfabetismo ?analfabetismo .
?estado dbpprop:nome ?nome .
?voto <http://ligadonospoliticos.com.br/politicobr/politician> ?politician .
?voto <http://ligadonospoliticos.com.br/politicobr/voting> ?votacao .
?votacao <http://ligadonospoliticos.com.br/politicobr/proposition> ?proposicao .
?proposicao rdf:type polbr:proposicao.
?proposicao dcterms:description ?ementa
FILTER (?analfabetismo > 20)
FILTER regex(?ementa, "educação","i")
}
http://pt.dbpedia.org/?q=exemplos
16. Linking Citizens and Politicians
● Connecting politicians to public opinion through
social media.
17. Conclusion
● Following best practices (such as LOD) to open data
allows citizen developers to create interesting apps
● We hope that more data and more applications will
engage citizens in an active dialog with the
government
● As future work, we intend to:
● Update and link more data.
● Provide a friendly interface for query formulation
● Link propositions to LOD via entity extraction
● Engage more developers
18. References
● UNITED NATIONS. “United Nations E-Government Survey 2010”. New Yourk: UN
Publishing Section, 2010,
“http://unpan1.un.org/intradoc/groups/public/documents/un /unpan038851.pdf”,
December 2010.
● Opinião Pública.: Vol. 15, no 2, November, 2009, Encarte Tendências, Campinas,
Brazil. p.511-542. Latin American Public Opinion (LAPOP). 2007.
http://www.scielo.br/pdf/op/v15n2/09.pdf. March, 2011.