Mapping Disinformation Campaigns across Platforms: The German General Election

Technical University of Munich

Abstract

During the German general election in 2017, there were coordinated attempts to disturb online public opinion. Disinformation campaigns used different social media platforms like Facebook and Twitter. They managed to infiltrate the online reporting of major German news outlets. Due to the distributed character of disinformation campaigns, these attempts were hardly noticed by the public, and their real dimension has not yet been revealed. We will use data from Facebook about shared URLs in combination with our unique dataset of the political online discourse in Germany, including 700 million tweets and 1.8 million URLs of media content shared by Twitter users, time series of media reports on the German political parties, and political polls to answer two questions: What was the real dimension of a disinformation campaign that was linked (by Twitter) to the Russian Internet Research Agency (IRA)? Is there a measurable effect of the disinformation campaign, either on news reports on the parties or directly on polls? To answer these questions, data access to the Facebook URL Shares dataset is requested to measure the distribution of content from disinformation campaigns across platforms. The starting point of the project is a list of Twitter accounts that Twitter has publicly linked to the IRA.

Research Team

Principal Investigator

Simon Hegelich

Professor of Political Data Science, Technical University of Munich

  • Bio ▾

    Simon Hegelich is professor of political data science at the Technical University of Munich. He works at the intersection of political science and computer science. He is interested in the political dimension of the ongoing digital revolution as well as in implementing new methods in political science such as machine learning. Hegelich studied political science at the University of Münster (Germany), where he also received his doctorate and acquired his postdoctoral teaching qualification (habilitation). From 2011 to 2016, he was managing director of the interdisciplinary research center FoKoS of the University of Siegen. In 2016, Hegelich was appointed associate professor at the Bavarian School of Public Policy. Since 2018, Hegelich has been appointed member of the Department of Computer Science at TUM. Hegelich has been advisor to the German government and the parliament on different topics of disinformation and the public online discourse.

Participants

Joana Bayraktar

Student Assistant, Technical University of Munich

  • Bio ▾

    Joana Bayraktar studies political science at the Bavarian School for Public Policy at the Technical University of Munich. Since August 2018, she has been a student assistant for the professorship for political data science.

Fabienne Marco

Research Assistant, Technical University of Munich

  • Bio ▾

    Fabienne Marco is studying politics and technology (MSc) and mathematics in data science (MSc) at the Technical University of Munich after completing her BSc in mathematics. In her master’s thesis she is dealing with classifying sexism in social media. She also passed different language certificates in Russian, English, and Spanish. During her studies, she gained experience at different companies as a working student. Since April 2017, she has worked for the professorship for political data science as student research assistant.

Orestis Papakyriakopoulos

Research Assistant, Technical University of Munich

  • Bio ▾

    Orestis Papakyriakopoulos is a researcher at the Bavarian School for Public Policy at the Technical University of Munich. He studied civil engineering (Dipl. Ing) at the National Technical University of Athens and philosophy of science and technology (MA) at the Technical University of Munich. In the past, he has served as an engineer, as well as in the communication and administration of science. Furthermore, he conducted research in civil engineering, risk analysis, and econometrics in Germany and the USA. Currently, Papakyriakopoulos is a research associate in political data science, analyzing questions of applied and philosophical nature on politics, digitalization, and artificial intelligence. Besides publishing scientific articles in the aforementioned fields, he has given various interviews. Papakyriakopoulos teaches theoretical and applied courses for bachelor’s and master’s students related to the field of political data science.

Juan Carlos Medina Serrano

Research Assistant, Technical University of Munich

  • Bio ▾

    Juan Carlos Medina Serrano is a data scientist working at the Bavarian School for Public Policy of the Technical University of Munich. He was born in Mexico and studied engineering physics at the Tecnológico de Monterrey. He later moved to Germany, where he obtained two master’s degrees: computational science and data engineering and analytics. Medina Serrano has worked as a data scientist for companies like BMW, Bosch, and Siemens. Currently, he is pursuing a PhD and his research involves the spread of misinformation in social media and its effects on the political landscape. He has participated in several radio interviews and in TV program such as The Stream from Al Jazeera and Sat.1.

Morteza Shahrezaye

Research Assistant, Technical University of Munich

  • Bio ▾

    Morteza Shahrezaye did his bachelor studies in economics and statistics in Iran. Afterwards he moved to Siegen, Germany, to do a master’s in economics. Due to his expertise in scientific programming, he was collaborating with three different research groups at university. He learned to develop mathematical and statistical models in different programming platforms. This helped him to start working at a German startup as machine learning developer, designing computer vision algorithms applied to panaromic videos. In March 2016, he moved to Munich to start his PhD as a computer scientist at Technical University of Munich (TUM). Shahrezaye is currently doing research mainly on networks and their use in social sciences. He specifically tries to extract knowledge from huge social activity networks consisting of billions of nodes and edges. The second branch of his research is natural language processing using artificial intelligence and deep learning tools. He also teaches statistics and mathematics at TUM.

Back to top