Mediating machines
Opportunities and Challenges of Artificial Intelligence (AI) in Peacemaking
About the project
In today’s technologized world, efforts to end armed conflict through mediation are increasingly shaped by digital technologies. While peacemaking traditionally has been considered a human-centered activity, negotiations are often supplemented with online activities. Social media has come to change the landscape of conflict, with conflict parties and stakeholders pro-actively using different online platforms to make their voices heard. The digitization of peace processes produces large amounts of data, which cannot be sufficiently analyzed using conventional means.
Artificial Intelligence (AI) offers opportunities to develop tools that could support mediators and their teams in gathering and analyzing the data produced in the context of peace processes. Although the topic of AI has come to receive increased attention in the field of peace mediation, there is a lack of concrete examples and applications that could bring discussions forward. This project will therefore explore the ways in which AI-applications could be used in peacemaking, while raising important questions about potential challenges these applications might pose.
In particular, the research will analyze how AI can support the analysis of arguments and opinions voiced by conflict parties and stakeholders in digital data. To this end, the project builds on recent advances in text-mining of online participation processes.
This project is conducted by researchers at the Centre on Conflict, Development and Peacebuilding (CCDP), Graduate Institute of International Development Studies, Geneva (IHEID), the Karlsruhe Institute of Technology (KIT), the Fraunhofer Institute of Optronics, System Technologies and Image Exploitation (IOSB), Universität Düsseldorf and SwissPeace.
The project is funded by the Artificial Intelligence and the Society of the Future Initiative of the Volkswagen Foundation.
Project activities
Workstream 1
Mediators' Needs and Ecosystems
The team will analyze the practical needs and potential strategic purposes of AI in mediation support, as well as the organizational and technological ecosystems in which AI tools could be embedded. We will examine existing efforts to create AI applications for mediators and explore how argument data could be used by mediators to assess dynamics of negotiation.
Workstream 2
Data and Methods
The team will discuss the methodological and data requirements of AI-supported argument analysis. We will look at options for transparent and explainable AI and related questions of inference methods. The team will also identify and analyze various data sources for machine-supported argument analysis, including data derived from social media and other digital inclusion efforts.
Workstream 3
Output Visualisation
The team will identify data visualization techniques that can help integrate AI outputs into mediators' workflows and decision-making processes. This includes geographic visualization and graph-based techniques graph-based techniques. We will also explore the prospects of creating visual data mining tools that present multiple perspectives on the data through several visualization styles and interactive data analysis.
Project updates
19 August 2020
Snapshots from our pilot analysis: What’s behind conflict stakeholders’ positions?
When leveraging the potential of machine learning (ML) tools for peace mediation, we are particularly interested in applications that enable meaningful analysis of unstructured text-data. Many available applications – including sentiment analysis – have their origins in for-profit use cases, and particularly, the advertisement industry. These tools may be well suited for automated tasks of…
Keep Reading →06 July 2020
Designing Digital Inclusion Project Report Published
The final report of the Designing Digital Inclusion Project, conducted in 2019 at the Graduate Institute, has now been published. While the report does not explicitly discuss Artificial Intelligence, it provides important background information on the emerging pratice of digital inclusion in the peace mediation field. Have a a look here.
Keep Reading →17 June 2020
Why Machine-supported Analysis Matters for Digital Inclusion
Until recently, mediators tended to think about their work as a "human-centered" activity. Facilitating negotiations between conflict parties was something that would occur predominantly offline – ideally, behind closed doors and in a technology-free environment. The 2020 global COVID-19 pandemic has radically changed this view. Due to lockdowns and global travel restrictions, many mediators and…
Keep Reading →11 May 2020
Arguing for Peace? What Argument Analysis Can Offer to Peace Mediation
The peace mediation literature seems to have relatively little to say about arguments. A quick google scholar search with keywords such as “mediation” and “arguments” or “arguing” does not yield a single match that would be relevant to the mediation of armed conflicts. Research on mediation seems more concerned with understanding factors that could impact mediation efforts, such as conflict type…
Keep Reading →20 April 2020
Share your insights!
The research team is currently conducting a needs assessment among mediators and mediation experts. Please get in touch if you would like to provide your insights and ideas!
Keep Reading →20 April 2020
CCDP joins the Cybermediation Network
In April 2020, the Centre on Conflict Development and Peacebuilding (CCDP) joined the Cybermediation Network. The CDDP will be represented by Andreas Hirblinger. The Network succeeds the CyberMediationInitiative, which comprised the United Nations Department of Political and Peacebuilding Affairs, the Centre for Humanitarian Dialogue (HD), DiploFoundation, and swisspeace. In addition to these…
Keep Reading →20 April 2020
Project website launched!!
We are happy to see our project website online! In the next weeks and months, we will use this space to share important updates and insights, as well as announce opportunities for collaboration!
Keep Reading →Follow us on twitter
Research team
Andreas Hirblinger
Dr. Andreas Hirblinger has broad expertise on inclusive peacemaking and peacebuilding. He conducts research on knowledge production, digital inclusion and the use of technology in peace processes at the CCDP, Graduate Institute, Geneva.
Laurent Goetschel
Prof. Laurent Goetschel is the Director of swisspeace, and the director of the Basel Peace Forum, which strives to bring the innovation and peacebuilding fields closer. He is also a professor of political science at the University of Basel.
Ahmed Eleiba
Ahmed Eleiba is the senior program officer for Mediation at swisspeace. He also represents swisspeace in the CyberMediation initiative, which focuses on the role of social media and digital technologies in peace mediation.
Tamara Mchedlidze
Dr. Tamara Mchedlidze is an expert in graph algorithms and network visualization, and has constructed visual representations of online participatory processes.
Gregor Betz
Prof. Gregor Betz specializes in reason and argumentation theory, opinion and argument maps, in addition to the application of digital technologies in argumentation.
David Lanius
Dr. David Lanius’ areas of expertise include argumentation analysis, argument theory and argumentation reconstruction. He is particularly interested in analyzing the use of alternative facts or “fake news” in populist debates.
Alexey Pak
Dr. Alexey Pak is an established researcher with a background in theoretical physics. His areas of expertise include practical machine learning techniques, including ANNs and Bayesian models.
Stefan Conrad
Prof. Stefan Conrad has expertise in machine learning and natural language processing, with a particular focus on argument mining of online participatory processes.
Michael Färber
Dr. Michael Färber’s research expertise includes natural language processing, machine learning and semantic web. His current focus is on scholarly data mining, scholarly recommender systems and knowledge graphs.
Background readings
Designing Digital Inclusion in Peacemaking
Arguments can be collected and analyzed as part of inclusive peacemaking efforts. Digital technologies play a key role in facilitating such processes. The most important insights from the "Designing Digital Inclusion" Project, conducted at the Graduate Institute of International Development Studies. The full text is available here.
Mediation and Artificial Intelligence:Notes on the Future of International Conflict Resolution
Artificial Intelligence plays an increasing role in Cybermediation and Digital Peacemaking. This report by the DiploFoundation provides an overview of the current debate on the use of Artificial Intelligence in Peace Mediation. The full text is available here.
Argument Mining: A Machine Learning Perspective
Understanding arguments often requires the analysis of large amounts of text data. This paper dicusses options, problems and challenges for machine-learning supported argument mining. The full text is available here.
Applying argumentation to structure and visualize multi-dimensional opinion spaces
Visualization is critical for understanding arguments. This paper discusses options to visualize large scale, multi-dimensional opinion spaces, based on formal argumentation theory. The full text is available here.
Text Mining für Online-Partizipationsverfahren: Die Notwendigkeit einer maschinell unterstützten Auswertung
Text mining and argument mining have been sucessfully explored in other participatory political processes, for instance in the context of large infrastructure projects. This article summarises important lessons learned from Germany. The full text is available here.
Related initiatives
UN Peacemaker Digital Toolkit
Online resource on digital technologies, provided by the United Nations Mediation Support Unit (MSU).
Digital Inclusion in Peacemaking
Research project conducted at the Centre on Conflict, Development and Peacebuilding (CCDP), Graduate Institute of International and Development Studies, 2020.