Tokyo, Japan — June 2026 — The Media Bias Group is pleased to announce that its proposal for a Dagstuhl Seminar has been accepted by Schloss Dagstuhl – Leibniz Center for Informatics, one of the world’s most prestigious venues for scientific exchange in computer science.
The seminar, entitled “Towards Standardizing Media Bias for Computational Research,” will bring together internationally leading researchers, journalists, policymakers, and AI industry experts to address one of the central challenges in computational media bias research: establishing a shared conceptual framework and taxonomy for the field.
Over the past decade, research on media bias has grown rapidly across natural language processing, computational social science, journalism, political science, and artificial intelligence. While considerable methodological progress has been made, the field still lacks broadly accepted standards for defining, categorizing, annotating, and evaluating different forms of media bias. As a result, comparing findings across studies and developing shared benchmarks remains challenging.
The Dagstuhl Seminar aims to address this gap by fostering interdisciplinary discussions on the conceptual foundations of media bias, annotation methodologies, benchmark design, computational evaluation, and future research priorities. Participants will work collaboratively throughout the week to refine a shared taxonomy that can serve as a foundation for future computational research, datasets, and evaluation frameworks.
The seminar is organized by Dr. Timo Spinde (National Institute of Informatics / University of Göttingen), Gunda Ehmke (Data Innovation Lab, German Federal Foreign Office), Prof. Dr. Gianluca Demartini (The University of Queensland), Prof. Dr. Isao Echizen (National Institute of Informatics), and Prof. Dr. Karsten Donnay (University of Zurich).
The event also represents another important milestone for the Media Bias Group. Since its foundation, MBG has worked to connect researchers across disciplines and countries to advance trustworthy and reproducible media bias research. Hosting a Dagstuhl Seminar provides a unique opportunity to bring together many of the leading voices in the field to jointly shape its future.
Beyond the seminar itself, the organizers plan to publish a refined and openly accessible media bias taxonomy, a comprehensive seminar report, and a roadmap for future collaboration, benchmark development, and community activities. The outcomes are intended to benefit researchers, journalists, policymakers, and AI practitioners working on trustworthy information systems and computational media analysis.
The Media Bias Group looks forward to welcoming participants from academia, industry, journalism, and policy for what promises to become a landmark event for the international media bias research community.
About Schloss Dagstuhl
Schloss Dagstuhl – Leibniz Center for Informatics is internationally recognized as one of the world’s leading venues for scientific workshops in computer science. Every year, only a limited number of seminar proposals are selected through a highly competitive international review process. Dagstuhl Seminars have become an important catalyst for shaping emerging research fields, fostering interdisciplinary collaboration, and initiating long-term international research partnerships.
About the Media Bias Group
The Media Bias Group is an international research initiative dedicated to advancing computational media bias analysis, trustworthy AI, and interdisciplinary research on information quality. The group brings together researchers from computer science, journalism, political science, psychology, and related disciplines to improve the understanding, measurement, and communication of media bias through open science and international collaboration.
