Visual Analytics - Past, Present, and Future

The Department of Homeland Security determined early in its history that visual analytics would be a key component in its ability to understand the threats posed by terrorism and the consequences, both immediate and long-term, of natural or manmade disasters. That broad need to ensure the security of our homeland through the application of novel technologies and techniques was the main reason that DHS decided to establish a National Visualization and Analytics Center in 2004. Since then, the NVAC effort has been supplemented the development of the associated complex of university centers, other government centers, and industry partnerships. Together, all of these elements constitute a globally based science and technology enterprise as well as an enduring talent base that will enable the US to not only address homeland security issues but also continue playing a leadership role in the field of visual analytics.

But the question remains, “Why visual analytics?” Many years ago, many of you began your involvement with the field of information analytics from the simple yet compelling viewpoint of knowledge management. The metaphors then, as now, were not instructive aids. Rather, they were worrisome specters: massive information flows, data deluge, “overload”, connecting the dots, and mining for “nuggets”. Knowledge discovery was the key, and presenting the greatest amount of information to the harried or bewildered user was the challenge. Visualization, and specifically visual analytics, was then seen as the next best hope to address that challenge. The strides made in the last 5 years have been great, and the progress phenomenal.

Nonetheless, we cannot be complacent and comfortable with the progress that has been made in visualization science and technology and its outstanding ability to address the problems of information “overload” and making “connections” and finding “nuggets”. As this talk will argue, capabilities in these areas must be replaced with new techniques based on cognitive principles and recent evolutionary findings. Intelligence analysis must make way for information or knowledge synthesis, retrospection must be replaced with prospective analysis, and understanding of static data must give way to real-time awareness of dynamic, wide-ranging data. The objective is ultimately to seek ways to deliver “designer information” – up-to-date, customized or tailored knowledge delivered just as needed. Visualization is the key to this quest, and we must all work together to discover and develop new visually based techniques that provide those capabilities.