Humanitarian Vis

This online meetup session features 2 talks. The Zoom passcode is Pd2ZW59n.

TALK 1: What do we know about humanitarian visualizations? From evoking compassion to promoting human welfare

Luiz Morais Postdoctoral Researcher Inria Bordeaux, France https://luizaugustomm.github.io

Humanitarian issues such as extreme poverty or refugee crises have been acknowledged worldwide over the past few decades. Organizations often have campaigns that focus on emotional strategies, such as showing the picture of a starving child or dead animals to persuade people to act. Another common tactic used by news media is reporting the issues through data visualizations. The psychology literature has extensively investigated the former approach, but there is still a lack of research on the effect of data visualizations to evoke emotions such as compassion or prosocial behaviors such as donating. This presentation shows recent contributions in that direction and discusses future perspectives of research.

---

TALK 2: Challenges of Uncertainty Visualization in Practice: First Insights from a Study with Epidemiologists

Christoph Kinkledey Professor of Media Informatics and Data Visualization HAW Hamburg https://kinkeldey.com

Today, there is a consensus that including information about data uncertainty can support data analysis processes and can lead to better informed insights from the data. Yet, people find it difficult to incorporate uncertainty into their analyses in a helpful way. Visual communication of uncertainty is a way to counter this challenge, but until today, most data analysis tools and workflows do not show uncertainty in the data. In this talk, Christoph presents typical challenges that prevent domain experts from using uncertainty visualization for data analysis in practice. First insights from a current joint project with the gicentre, City, University of London illustrate new ways to better support domain experts in tackling these challenges. The project involves epidemiologists from Australia in online meetings to capture their current strategies for dealing with uncertainty in epidemiological data analysis. It further explores potential extended strategies using visual prototypes designed specifically to support real analytical tasks from their work. Christoph talks about the goal to create guidelines for designing uncertainty visualizations that actually help domain experts with their data analyses.

People involved

• Alexis Pister
• Uta Hinrichs (Google Scholar)
• Benjamin Bach (Google Scholar)
• Xinhuan Shu