Journalism plays a critical role in democratic societies by providing citizens with reliable information, grounded in an ethical commitment to accuracy, verification, and fairness, as the Society of Professional Journalists’ Code of Ethics reminds us, “Ethical journalism should be accurate and fair.” Yet, even as data‐driven storytelling reshapes newsrooms around the world, adoption remains uneven. Despite Japan’s celebrated design heritage, its major outlets have been comparatively slow to embrace advanced data visualizations: Yomiuri Shimbun—one of the world’s highest-circulation newspapers—only recently launched a data journalism training program in partnership with Emerson College, underscoring a broader lag in Japanese print media.

At the same time, the advent of generative AI brings both promise and peril. On one hand, AI could automate aspects of accessible figure generation and create personalized data narratives; on the other, it poses fresh challenges to journalistic deontology, from safeguarding factual integrity to preventing misleading “hallucinations.” We therefore need a focused research agenda that not only maps the current accessibility gap but also prototypes AI-augmented workflows which uphold ethical standards while broadening access to data-driven stories.

During this seminar, we will bring together journalists, visualization experts, and accessibility researchers to foster rich discussions around the evolving role of visualization in journalism. Our goal is to identify and expand real-world application scenarios, exploring how AI-driven interaction techniques, multimodal interfaces, and immersive media might transform data journalism over the next decade, and to co-create actionable guidelines and prototype tools that ensure inclusive, ethical, and innovative storytelling.

Journalism plays a critical role in democratic societies by providing citizens with reliable information. At its core is an ethical commitment to accuracy, verification, and fairness, as articulated in the Society of Professional Journalists’ Code of Ethics: “Ethical journalism should be accurate and fair” [1]. Journalists must rigorously fact-check and corroborate their material to uphold public trust and fulfill their professional deontology. Audiences today engage with news through a variety of modalities. While traditional reporting was once confined to text, modern journalism embraces multimedia—audio podcasts, video segments, interactive articles—and especially infographics and data visualizations. These visual formats have become indispensable for conveying complex datasets in an engaging, digestible manner, a shift detailed in Data-Driven Storytelling [2].

Over the past decade, data-driven storytelling has proliferated among leading outlets. During the Olympic Games, for example, The New York Times (https://www.nytimes.com/) produced interactive graphics illustrating athletes’ performance trajectories against historical benchmarks [3]. Similarly, in recent U.S. presidential elections, news organizations have offered real-time visual narratives of vote counts methods analyzed in Johnson and Lee’s comparative study of election visualization techniques [4]. Academic and industry authorities now routinely integrate visual content into their production pipelines not merely to attract attention, but to provide multiple analytical perspectives that reinforce editorial messages [5]. This practice is widespread in Western media markets (e.g., the United States and Europe). However, despite Japan’s design heritage, its major news outlets have been comparatively slow to adopt advanced data visualizations. Yomiuri Shimbun—one of the world’s highest-circulation newspapers—only recently launched a data journalism training program in partnership with Emerson College, highlighting a broader lag in Japanese print media [6].

While interest in data visualization within journalistic content has grown markedly, substantial research and development remain to be done. First, the visual production of data stories is neither standardized nor fully understood. There is tremendous variability in form and design, which can be beneficial for diverse communication goals, but current practice is heavily skewed toward maps, a narrow subset of the rich spectrum of visualization techniques available [9]. Understanding the reasons behind this mapping bias and developing a roadmap for broader democratization and data‐literacy training, is therefore essential [10]. Second, the availability of data‐driven content remains uneven for users with disabilities: infographics and interactive charts frequently lack meaningful alt text, screen‐reader compatibility, or adaptations for those with visual or cognitive impairments [7]. Without deliberate accessible design, data‐rich journalism risks excluding audiences who most need clear, reliable information. Finally, with the advent of generative AI, there is both opportunity and risk: AI could automate aspects of accessible figure generation or personalized data narratives, but it also poses challenges to journalistic deontology—such as ensuring factual integrity and avoiding misleading “hallucinations.” A focused research agenda must therefore not only characterize the current accessibility gap but also prototype AI‐augmented workflows that uphold ethical standards while expanding access to data‐driven stories [11] in the journalism domain.

This seminar arrives at a pivotal moment: regulatory frameworks like the European Web Accessibility Directive are compelling digital publishers to ensure compliance with accessibility standards [8], and growing awareness of information equity has heightened demand for inclusive journalistic content. By convening journalists, visualization experts, and accessibility researchers, our seminar will harness this momentum to develop actionable guidelines and prototype tools that embed universal access into the data-visualization pipeline—ensuring data-driven stories engage and inform all audience segments.

During this seminar, we will foster discussions with researchers and practitioners around visualization in journalism material. The proposed seminar is related to the following list of previous Shonan and Dagstuhl meetings:

• NII Shonan Meeting No. 64: Big Data Visual Analytics (Nov 8–11 2015)Link
• NII Shonan Meeting No. 213: Augmented Multimodal Interaction for Synchronous Presentation, Collaboration, and Education with Remote Audiences (Jun 24–27 2024) Link.
• Dagstuhl Seminar 23252: Inclusive Data Visualization (Jun 18–23 2023) Link
• Dagstuhl Seminar 24301: Art, Visual Illusions, and Data Visualization (Jul 21–26 2024) Link.

Our seminar will serve as a forum to identify and expand application scenarios for journalistic visualization, envisioning how interaction techniques driven by AI, multimodal interfaces, and immersive media might evolve over the next decade. We anticipate that the meeting’s outcomes will resonate across multiple academic communities, including those affiliated with ACM SIGCHI (CHI, CSCW, UIST) and the IEEE VGTC (VIS, PacificVis, EuroVis). Throughout the seminar, participants will grapple with several core research questions. First, how can AI enhance accessibility in data‐rich journalism, and what are its current limitations? Second, under what conditions can AI‐driven visualization tools be deemed trustworthy and aligned with journalistic ethics? Third, why do many news graphics default to map‐based representations, and how can practitioners move beyond them to embrace a broader repertoire of visualization techniques? Finally, what emerging challenges technological, social, or regulatory will shape the future of data journalism? Ultimately, our goal is to articulate timely, urgent, and actionable research directions that participants can pursue after our Shonan seminar. We will synthesize these insights into vision papers and position statements aimed at premier venues such as ACM CHI and IEEE TVCG. The intensive, collaborative format of the meeting will enable us to generate outcomes tailored to both specialist and general audiences, which we will share with the wider community via a dedicated “Seminar Outcomes” portal. As a capstone, we will package our outputs into shareable resources: recorded talks, interactive demonstrations, pedagogical guides, and a state-of-the-art compendium providing an inspirational toolkit for researchers, designers, data journalists, and newsrooms alike.

Day 1 – Foundations & Shared Vocabulary

• Opening keynote by the organizers (Hurter, Hurtut, Itoh)
• Invited keynote by a leading researcher in accessibility or journalism
• Plenary talks: multimodal storytelling, accessibility challenges in data journalism
• Breakout Session A: Identifying core accessibility barriers

Day 2 – Tools, Methods & Case Studies

• Tutorials: automated alt-text, data sonification, storytelling workflows
• Lightning talks by participants
• Breakout Session B: scenario-based design
• Panel discussion: The future of accessible computational journalism

Day 3 – Design Sprint & Knowledge Consolidation

• Collaborative design sprint
• Cross-group critique session
• Half-day excursion to be defined

Follow-Up – Outcomes, Writing & Roadmap

• Dedicated writing blocks for vision and position papers
• Editorial meeting to define concrete outputs (publication, review paper, potential Open Access book)
• Definition of a 6-month roadmap and post-seminar collaborations

[1] Society of Professional Journalists. (2014). SPJ Code of Ethics. Retrieved from https://www.spj.org/ethicscode.asp

[2] Riche, N. H., Hurter, C., Diakopoulos, N., & Carpendale, S. (2018). Data-Driven Storytelling. CRC Press.

[3] The New York Times. (2021). How Olympic Athletes Compare Across History. The New York Times.

[4] Johnson, A., & Lee, B. (2024). Visualizing Election Results: A Comparative Study. In Proceedings of the CHI Conference on Human Factors in Computing Systems.

[5] Hullman, A., & Diakopoulos, N. (2011). Visualization rhetoric: Framing effects in narrative visualization. IEEE Transactions on Visualization and Computer Graphics, 17(12), 2231–2240.

[6] Yomiuri Shimbun. (2023). Yomiuri Shimbun Partners with Emerson College to Launch Data Journalism Training Program. Yomiuri Shimbun.

[7] Slayton, K. (2020). Accessibility in Data Visualization: A Survey. Journal of Visual Communication and Image Representation, 75, 102978.

[8] European Commission. (2021). Report on the Implementation of the Web Accessibility Directive. Publications Office of the European Union.

[9] Heer, J., Bostock, M., & Ogievetsky, V. (2010). A Tour through the Visualization Zoo. Communications of the ACM, 53(6), 59–67.

[10] Lee, B., Kazi, R. H., & Smith, G. A. (2015). Designing with Data: Improving Accessibility and Literacy Through Visualization Workshops. IEEE VIS Workshop on Visualization for the Digital Humanities.

[11] McMillan, B., & McCoy, J. (2024). Generative AI in Newsrooms: Ethical Implications for Automated Visualization. Journalism Practice, 18(2), 234–251.