Skip to content

Quadrilingual corpora

A data bank of political and media discourses about Russia's invasion of Ukraine in five countries, two belligerents, the United States, the United Kingdom, and France, includes war-related

  1. speeches of political leaders (Volodymyr Zelensky, Vladimir Putin, Joe Biden, Emmanuel Macron and the Prime Ministers of the United Kingdom Boris Johnson, Liz Truss, Rishi Sunak and Keir Starmer),
  2. debates in national legislatures (Ukrainian Rada, Russian Duma, U.S. Congress, British Parliament and French Assemblée nationale),
  3. news items published in legacy media (ICTV, RBC Ukraina, Kommersant, Izvestia, First TV Channel, the New York Times, the Washington Post, USA Today, Fox, the Times, and Le Monde),
  4. news items published in digital media (Ukrainska Pravda, Liga, Strana, Gazeta.ru, Meduza), and
  5. posts in social media (VKontakte, Telegram).

The data bank offers continuous coverage of the first three years of the all-out war (January 2022 to February 2025). It contains 341 million words in four languages: Ukrainian, Russian, English and French.

 

Dictionary of war

A quadrilingual "dictionary of war" has been created to mine and analyze big textual data about the invasion. It includes 600+ categories from "adversary" to "Zelensky." Each category contains several words and n-grams.

 

Patterns in political and media discourses

With the help of a dictionary-assisted analysis, it is possible to identify and study changing patterns in war coverage. For instance, the figure below visualizes similarities between various sources of political and media discourses about Russia’s invasion of Ukraine during the first three years of the war (Stress I = ).154).

 

A chart visualizing dynamics of the frequencies of mentions of US President Donald Trump in news reports about the war during the same period serve as the other illustration (the US Presidential election was held on Day 986*, the inauguration was held on Day 1061*). More examples, and a description of the original methodology, can be found in the recently published scholarly articles and a monograph.