Research on agentic visualization receives IEEE Best Paper Award runner-up recognition
The paper Agentic Visualization: Extracting Agent-Based Design Patterns From Visualization Systems has been selected as runner-up for the 2025 Best Paper Award by IEEE Computer Graphics and Applications (IEEE CG&A), recognising outstanding research published in the journal over the past year.
The paper was authored by Vaishali Dhanoa, Anton Wolter, Gabriela Molina León, Hans-Jörg Schulz and Niklas Elmqvist. The annual award programme celebrates research excellence, innovation and impact across the IEEE Computer Society's publications, with each publication selecting one Best Paper Award recipient and, optionally, one runner-up.
The research introduces the concept of agentic visualization, interactive visual analysis systems that incorporate autonomous agents while preserving human analytical control. As AI agents become increasingly capable of reasoning and acting autonomously, the paper explores how visualization systems can support effective collaboration between humans and AI during data analysis and decision-making.
Rather than viewing human and AI agency as competing forces, the researchers propose a framework in which both can contribute their respective strengths. Agentic visualization enables AI agents to independently perform complex analytical tasks while allowing human users to guide the analytical process, interpret results and maintain decision-making authority.
By analysing existing visualization systems that include autonomous or semi-autonomous AI components, the researchers identified a collection of reusable design patterns that can guide the development of future visualization tools. These patterns describe how AI agents can take on different roles, communicate findings to users and coordinate analytical tasks while maintaining transparency and human oversight.
The paper argues that visualization will play an increasingly important role in helping people understand and collaborate with AI agents. Its proposed design patterns provide a foundation for developing human-centred AI systems that expand analytical capabilities while preserving the human insight and interpretability that have long been central to visualization research.
Read ‘Agentic Visualization: Extracting Agent-Based Design Patterns From Visualization Systems’.