Time: Wednesday, October 16, 2024 12:00 PM-1:15 PM EDT (UTC-4)

In 2004, Visual Analytics emerged as an outgrowth of scientific visualization and information visualization with a focus on the mechanisms of analytic reasoning facilitated by interactive visual interfaces. Now, twenty years later, this panel invites pioneers from the field of visual analytics to talk about the foundations of visual analytics, key works that helped define the field, and what they see as the emerging challenges for the next generation of visual analytics researchers.

With the founding of the National Visual Analytics Center at Pacific Northwest National Laboratory in 2004 and the publication of Illuminating the Path in 2005, visual analytics emerged as "the science of analytical reasoning facilitated by interactive visual interfaces." The Symposium on Visual Analytics Science and Technology (VAST) was established soon thereafter (2006), becoming part of the main IEEE Visualization Week conference in 2009. The VAST Challenge, first held in 2006, aims to advance visual analytics through competition by engaging participants from industry, government, and academia with realistic tasks and data sets, fostering innovation in data transformations, visualizations, and analytic techniques. In 2008, the VisMaster initiative funded by the European Union started, publishing its Visual Analytics roadmap in 2010. Visualization Week (or VisWeek) would eventually become VIS (VAST-InfoVis-SciVis), and the three main conference tracks would be replaced with an area chair model in 2021. Throughout its history, the visual analytics community has developed and deployed technology (e.g., Jigsaw, SensePlace2), explored numerous analytical challenges from various domains (e.g., movement analysis [Andrienko et al., 2008], [Ferreira et al., 2013], text analysis [Collins et al., 2009], [Whiting and Cramer, 2002], dimension reduction, temporal analysis, patent analysis, syndromic surveillance, fraud detection, and more), and helped enable the world to detect the unexpected and discover the unexpected (e.g., [Kandel et al., 2012], [Weber et al., 2012]).

As such, Visual Analytics has become a central part of IEEE VIS and Keim's visual analytics mantra ("Analyze first, show the important, zoom, filter and analyze further, details on demand.") was astute in leaning into machine learning and artificial intelligence as a partner in the analytic and sensemaking process. Today, visual analytics research is embedded into machine learning to help in model building (e.g., [Endert et al., 2012], [Sacha et al., 2019]), diagnosis (e.g., [Guo et al., 2021], [Ming et al., 2017]), and verification (e.g., [Andrienko et al., 2008], [Ming et al., 2019]). Large Language Models are serving as conversational agents to support interactive data analysis [Narechania et al., 2021], and the cognitive aspects of Human-AI teaming are becoming critical in deploying technology [Padilla et al., 2023], [Zhao et al., 2024], [Wall et al., 2017]. In this panel, we invite some of the pioneers of visual analytics to share with the community the journey of this area, reflecting on what worked, what did not work, and how they see visual analytics research evolving over the next 20 years.

Panelists

Organizers