Linking the European Parliament debates to Italian Parliament data

Linking the European debates to the Italian Parliament data:

research opportunities and tools

Silvia Giannini – Politecnico di Bari (IT)

Talk of Europe - Travelling CLARIN Campus Hilversum, 6 – 10 October 2014

The Italian Parliament (IP)

• Legislative power

The Chamber of Deputies Linked Open Data

• 490.554.781 triples updated every day* • Description of biographical data, parliamentary

mandates, parliamentary bodies, laws and acts drafts, debates, votes, elections, etc. of all the XVII legislatures

• OWL ontology for domain specific Classes and predicates, FOAF and DC vocabularies for metadata description

• Public SPARQL endpoint (http://dati.camera.it/sparql)

• Senate LOD under development (http://dati.senato.it/)

*Source: http://dati.camera.it/it/ - English version: http://data.camera.it/data/en/ - 07.10.2014

Applications over IP open data

• Historical Portal and Institutional Atlas (http://data.camera.it/data/en/apps/)

• Code4Italy (http://dati.camera.it/it/hackathon/)

• Openpolis Association (http://openpolis.it/eng/)

Historical Portal and Institutional Atlas

Code4Italy - SocialCamera

http://opensocial.camera/

Code4Italy - PaLmaS

http://sinatra.cirsfid.unibo.it/code4italynew/

Code4Italy – Camera4DUMMIES

http://marcoslin.github.io/Code4Italy/#/home

openpolis

Openpolis - openparlamento

Openpolis - openparlamento

• Open Politicians: who our representatives, from the smallest Italian municipality to the European Parliament, are and what they are doing and saying

Openpolis - openparlamento

• Open Politicians: who our representatives, from the smallest Italian municipality to the European Parliament, are and what they are doing and saying

Openpolis - openpolitici

• You Are Here: by comparing your opinion on major political issues to those of the candidates, it helps you to find out who are the closest and farthest candidates to you in European, national and local elections

‘’Persona’’ dataset

• The registry of persons graph (persona.rdf) includes deputies, senators and non-parliamentary members of the government

ToE Speaker/IP Person Alignment

• 458 instances of Speaker representing the Italian country (EUCountry_IT)

• Alignment using foaf:givenName, foaf:familyName and lp:dateBirth properties

• 210 exact matches (47 DBpedia, 24 Italian Dbpedia, 13 Freebase links)

• 5 homonyms manually resolved • 243 missing links (a EP eligible deputy is every

Italian citizen older than 25 years)

• Every elected deputy joins a Parliamentary group that reflects its party ideas (it does not always coincide with the national party)

‘’Deputato’’ dataset

Instances Alignment (2/2)

• 59 different National Italian parties in the ToE dataset associated with Italian speakers

• 55 new links between 39 Italian speakers in the ToE dataset and the corresponding Italian Parliamentary groups they were member of.

• The information is available till 2001 (with lex 07/04/2004 it is prohibited to hold a national mandate together with a European one)

Information Integration

• DEMO (http://193.204.59.20/toe/fini.html)1

• Analyze non-attendace rate with numbers of concurring appointments2

1 Realized with http://timeline.knightlab.com/ 2http://www.votewatch.eu/blog/europarlamento-italia-brilla-per-assenteismo-alle-votazioni/

Clustering RDF datasets

• Algorithm based on Common Subsumers (CSs) extraction (the most specific description shared by two RDF resources)

S. Colucci, S. Giannini, F.M. Donini, E. Di Sciascio, Finding Commonalities in Linked Open Data, Proceedings of the 29th Italian Conference on Computational Logic (CILC 2014), Torino, Italy, June 2014, pp. 37 – 42

Clustering RDF datasets

Deputies Nilde Iotti (ocd:d3140_10) and Tina Anselmi (ocd:d270_10) of the 10th Legislature

Clustering RDF datasets

SELECT DISTINCT ?x0 WHERE{ ?x0 a <http://dati.camera.it/ocd/deputato> . ?x0 <http://dati.camera.it/ocd/rif_leg> <http://dati.camera.it/ocd/legislatura.rdf/repubblica_10> . ?x0 <http://dati.camera.it/ocd/rif_mandatoCamera> ?x1 . ?x0 <http://xmlns.com/foaf/0.1/gender> female . ?x0 <http://purl.org/dc/elements/1.1/description> Laurea in lettere; insegnante@it . }

Clustering RDF datasets

Clustering RDF datasets

Clusters of Deputies from the 1st Legislature

Enriching the ToE dataset

• Several AgendaItem instances have the same title over years

• Track the evolution of topics in debates related by the same title

• Triplifying speeches text • Tools: LDA package, TextRazor API • Output: - Topics for sets of Speech instances with same date and title - Wikipedia/DBpedia pages describing the extracted topics

In collaboration with Albert Meroño – VU Amsterdam (NL)