Abstracts

  1. Follow-up CC2
  2. Is My Politician Coherent? Linking Institutional Speeches and Public Discourse of MEPs
  3. Discovering Links in Political Conversations
  4. Debates of the European Parliament: Just European or Global?
  5. Chambers to Chambers: ECJ Rulings in European Parliamentary Debate
  6. Polish Political News Dataset

(These abstracts can be downloaded as a pdf file here)
______________________________________________________________

Follow-up CC2

Wim Peters (Natural Language Processing Group, Department of Computer Science, University of Sheffield, United Kingdom) and Maarten Marx (Informatics Institute, University of Amsterdam, the Netherlands).

Recent parliamentary proceedings (NL, UK) (available in XML and RDF through http://search.politicalmashup.nl) are linked to the Talk of Europe EU speeches set, making them interoperable at speech level. We plan to incorporate the EU speeches into the Politicalmashup application as well. Our intention is to use and extend the work performed by USFD in Creative Camp #2, and provide the following information:

1. Domain modelling in the form of named entity recognition, term extraction and conceptual organization;
2. Linked open data resolution of named entities (persons, locations, organizations);
3. Links between NL/GB terms derived from the EU proceedings;
4. Term clustering for semantic exploration and comparison of parliamentary vs media language use.

Our intended use case is transnational migration. Immigration is a topic widely studied (eg. within DiLiPaD, at UvA in the ASCoR group of Vliegenthart and de Vreese). The created dataset would facilitate comparative research across nation states. The outcome will predominantly be in the form of annotated triples consisting of a URI linking to (a part of) a debate, most often a speech a URI linking to an extracted entity, term, organization a labeled link between the two URI’s provenance information about the linking process (eg., a confidence value, date, used software, etc.) These triples will be made available in CSV format. The Talk of Europe EU dataset will be incorporated into the parliamentary debates repository Search.Politicalmashup.nl.

Is My Politician Coherent? Linking Institutional Speeches and Public Discourse of MEPs

Alessandro Piscopo, Luis-Daniel Ibáñez, Pavlos Vougiouklis, Nikolaos Liappas (Web and Internet Science Team, University of Southampton, United Kingdom).

The European Parliament (EP) is the institution in charge of the legislative function of the EU. To help citizens understand whether institutional acts of the EP members are consistent with their public discourse and whether they protect the interests of the constituency in which they were elected, we propose to:

1. Link MEP’s speeches in the EP to tweets and Facebook posts published on their social accounts to concepts, using techniques applied in previous creative camps;
2. Compute values of the sentiment speeches express towards a concept, using state of the art sentiment analysis techniques;
3. Link EP speeches to the geographical places mentioned in them, i.e. to IRIs in GeoNames, using state of the art entity recognition techniques.

These linkages would enable an assessment of the coherency between political action and public discourse of MEPs, by connecting their institutional behaviour and their social media declarations based on both the concepts covered and the places mentioned, according to the related sentiment. Results will be published as extensions to the LinkedEP dataset.

Discovering Links in Political Conversations

 Mostafa Dehghani, Hosein Azarbonyad and Alex Olieman (Exploratory Political Search Project, University of Amsterdam, the Netherlands).

Creating enriched annotated parliamentary proceedings by linking contained concepts and named entities to external resources allows investigating information in a straightforward way. According to the fact that political proceedings are considered as domain-specific texts (‘political’) with specific characteristics (‘conversation’), and since entity linking systems typically perform very differently for different data sets and domains, there is something to be gained by specializing entity recognition and disambiguation for parliamentary proceedings, using the particular features, structures and already existing information in this data. While general-purpose entity linking systems are usually able to link to a large number of entities from various domains, they are not necessarily able to link the most salient entities in a given domain, and may confuse salient entities with similarly named counterparts from popular culture. So, we propose an entity linker in which a specialized linker is combined with general entity linking systems to achieve a promising precision as well as an acceptable recall. Furthermore, in order to link detected entities to the particular external resources, we proposed ‘indirect linking’ in which links are created based on the conceptual similarity between documents in the linked corpus and document in the external resource. Our target is to create links in the European Parliamentary Debates by focusing on the Dutch language and specific Dutch entities.

Debates of the European Parliament: Just European or Global?

 Solveiga Inokaitytė (Kaunas University of Technology, Lithuania), Dainius Jocas (TokenMill, UAB, Lithuania), Justina Madravickaitė (Vilnius University and Baltic Institute of Advanced Technologies, Lithuania), Vytautas Mickevičius (Vytautas Magnus University and Baltic Institute of Advanced Technologies, Lithuania), Rūta Užupytė (Vilnius University and Baltic Institute of Advanced Technologies, Lithuania).

We propose to link debates of the European Parliament with the GDELT Event Database (http://gdeltproject.org). Functions of the European Parliament in the multilevel European governance are still rather limited. However, it is becoming more and more similar to the ‘usual’ – national – parliaments. Importantly, it debates issues related not only to European matters but also topics that reflect the global agenda of international politics. Therefore, different hypotheses might be formulated with regard to which issues are included into the agenda of the EP and why. However, in order to test such hypotheses researchers need linked data wherein events of global agenda are linked with debates of the EP. Linking debates of the European Parliament with the GDELT Event Database involves:

1. Linking events in diverse collections.
2. Events identification and classification using CAMEO scheme (used GDELT) in EP debates.
3. Visualization: discussed events, related factions and MPs, who says what, etc.

Chambers to Chambers: ECJ Rulings in European Parliamentary Debate

Karin van Leeuwen (European Studies, University of Amsterdam, the Netherlands), Hilde Reiding (Centre for Parliamentary History, Nijmegen University, the Netherlands), Bart Vredebregt (Artificial Intelligence, University of Amsterdam, the Netherlands) and Radboud Winkels (Computational Legal Theory, Faculty of Law, University of Amsterdam, the Netherlands).

In the early initiatives on digital government data, the data corpus produced by Montesquieu’s third power, the judiciary, has so far been left aside in spite of the rich digital sources it has produced. Yet, especially on the European level, the role of the judiciary has from the early years of integration onwards been highly political, as has been acknowledged in a growing body of research on European legal politics and the history of European law. One of the themes puzzling political scientists as well as historians in this newly discovered research area is the acceptance of this political role of courts, in particular the European Court of Justice (ECJ). While the legitimacy of its rulings may be self-evident from a legal perspective, political scientists have pointed out the important role of interaction of the ECJ with other supranational institutions and with crucial actors within the Member States. This project aims to sophisticate this explanation by tracing the response to ECJ rulings in other political arenas such as parliament (proceedings) and the public sphere (media). As a first step, the Creative Camp will be used to link the data of ECJ rulings with proceedings of the European Parliament (EP).

Polish Political News Dataset

Ewa Kowalczuk (Institute of Computing Science, Poznan University of Technology, Poland).

Since Poland joined the European Union the internal and European politics interweave. Not many citizens are aware how European Parliament work affects their daily life, existing regulations and state policies. By linking the Plenary Debates dataset to the Polish Political News dataset I would like to indicate the topics which are discussed both at local and the European level and also jointly present the activities of the MEPs in the Parliament as well as on the local political scene. The subjects of the plenary debates were summarised in Adams Funk’s and Wim Peters’s enrichment, developed during the second Creative Camp. Polish Political News dataset contains the subjects of the transcribed pieces of television news, marked with Wordnet synets and DBpedia concepts. By adding the same links to the subjects of the plenary debates, I will be able to connect the both datasets. The links will be created automatically, based on the string similarity and DBpedia categories, if applicable. They will be subsequently manually reviewed in order to ensure the high quality of outcome. The links connecting MEPs occurring in the news will be developed in the analogous manner. The combined information might be of assistance to the citizens willing to inspect their activity, for example during the electoral process.