About PoliMedia

What is PoliMedia?

How do media cover political debates? Answering this question requires a cross-media analysis of the minutes of the political debates, newspaper articles & photos and radio bulletins. However, this type of research is currently laborious. You will need to go to a physical archive, or you can work with a multitude of online archives with different interfaces and limitations.

The PoliMedia project linked the minutes of the debates in the Dutch Parliament (Dutch Hansard) to the databases of historical newspapers and ANP radio bulletins to allow cross-media analysis of coverage in a uniform search interface.

In 2013 PoliMedia won the Veni Competition of the LinkedUp Challenge and was finalist of the Semantic Web Challenge.

How does PoliMedia work?

For each fragment from a single speaker in a debate, we extracted relevant information: the speaker, the date, important terms from its content and important terms from the description of the complete debate. This information was then combined to create a query with which we searched the archives of newspapers, radio bulletins and television programmes. Media items that corresponded to this query were retrieved and a link was created between the speech and the media item, creating a Semantic Web of Dutch Hansard and media coverage. This Semantic Web contains links from the Dutch Hansard to newspaper articles and radio bulettins. From evaluations it was found that there was a 62% recall and 75% precision.

To navigate this Semantic Web, a search user interface was developed based on a requirements study with five scholars in history and political communication. We created a faceted search interface in which the Dutch parliamentary minutes can be searched in full-text and in which refinements can be performed based on the speaker, the role of the speaker (parliament of government), political party and year. These debates are presented with links to the original locations of the media items.

PoliMedia poster

Who created PoliMedia?

PoliMedia is a CLARIN-NL project. It is a collaboration of the TU Delft and the Free University (development of Semantic Web of Dutch Hansard and media), the Netherlands Institute of Sound and Vision (development of the search user interface) and Erasmus University Rotterdam (projectleader & user research of historians and political communication researchers).

Future research

PoliMedia also aimed to link the Dutch Hansard with television programs. However. within this one-year project the creation of those links proved unsuccesful. In the future we will experiment with technology to use subtitles or automatic speech recognition to broaden the scope of the archival search.

In the follow-up project Talk of Europe the debates of the European Parliament are made compatible for linking with other media, for more information see www.talkofeurope.eu.

Presentations

  • Connecting political data to media data. Presentation at ASCoR Spring Colloquium ‘Big Data at the University of Amsterdam’, 18 February 2014, Amsterdam, Netherlands

  • PoliMedia. IAMHIST. Leicester, UK, 18 juli 2013 slideshare

  • Building the PoliMedia search system; data- and user-driven. eHumanities Group. Amsterdam, 18 april 2013 slideshare

  • PoliMedia symposium: linking political debates and media. Amsterdam, 23 januari 2013

    1. Introduction to PoliMedia slideshare
    2. Linking the data sets slideshare
    3. User research for the development of PoliMedia slideshare

  • User research for the development of search systems. Erasmus Universiteit Bibliotheek, 11 december 2012, Rotterdam slideshare

  • Bringing parliamentary debates to the Semantic Web. Workshop on Detection, Representation, and Exploitation of Events in the Semantic Web (DeRiVE 2012). Boston, VS, 12 november 2012 slideshare

  • PoliMedia introduction. Nederlandse Organisatie voor Taal- en Spraaktechnologie. The Hague, 22 juni 2012 slideshare

  • PoliMedia introduction. CLARIN kick-off. Hilversum, 6 maart 2012 slideshare

Publications