What is Visual Analytics?
Visual analytics (VA) was initially proposed as a means to help United States intelligence analysts meet the challenge of dealing with the masses of security-related information made available to them following the terrorist attacks on September 11, 2001, on the World Trade Center and Pentagon. They literally were lost in a data deluge.
Visual analytics is defined as “the science of analytical reasoning facilitated by interactive visual interfaces.”
It is a multidisciplinary field intended to help people understand how to synthesize information in order to derive insights from massive, dynamic, ambiguous, and often conflicting data. In practice, it helps skilled analysts rapidly explore large, complex data sets to gain new insights using interactive visualizations. It draws upon research in a number of relevant areas, including information visualization, human computer interaction, machine learning, statistics, and cognitive science.
Scope of Visual Analytics 
Visual analytics proposes to take advantage of the visual intelligence and cognitive capabilities of human analysts using interactive, exploratory, visualisation tools. These tools also include statistical analysis and machine learning capabilities that would be guided by the analyst.
The amount of data being stored in digital databases is fast growing beyond the capabilities of even expert analysts to manage and use in a coherent manner. Many organizations are trying to find ways to deal with this growing data tsunami.
Although visual analytics originally was intended to solve data analysis problems in the security and intelligence domains, the tools and techniques that have been developed in the past 6-8 years are also of great interest to analysts in many other data-rich domains, e.g., aerospace safety, manufacturing and maintenance, transportation, financial risk analysis and fraud detection, business intelligence and process analysis, health care and medical research, and environmental health and safety.
A World Data Deluge
Organizations everywhere are encountering a significant challenge: How to derive more valuable information from datasets that continue to increase in complexity and size. Traditional analytic approaches are often inadequate to cope with such data sets; important—and sometimes critical—information can remain undiscovered. Consequently, new methods are needed to enable analysts to explore data on this increasing scale to allow confirmation of expected findings and discovery of unexpected findings. Many organizations, researchers and analysts have identified VA as a promising approach for critical data analytic needs.
Global digitally stored data was estimated to be 195 exabytes (1018) in 2007 and is expected to grow to 1.8 zettabytes (1021) by 2011 (for comparison, the US Library of Congress collection is estimated to contain approximately 10 petabytes (1015) of data).
Why Visual Analytics?
- Raw data has little intrinsic value.
- Data mining can help find expected patterns, e.g., prospect for gold and find gold in the data.
- Visual analytics will help analysts see and explore their data to not only find the expected, but also discover the unexpected, e.g., look for gold and find gold, but also possibly find silver or copper in the same data.
Humans have very impressive visual and cognitive capabilities, but humans change very slowly, e.g., brain volume has only doubled in approximately 2.5 x 106 years.
Computing technology, however, has been changing very quickly, e.g., Moore’s Law demonstrates that integrated circuit capacity has consistently doubled in approximately 2 years periods.
One goal of visual analytics is to build better tools and develop better methods to take advantage of human visual and cognitive problem solving capabilities.
 Thomas, J.J. & Cook, K.A. (Eds). Illuminating the path: The research and development agenda for visual analytics. IEEE Computer Society, 2005. (See pdf at NVAC).
 Keim, D.A. & Mansmann, F. Schneidewind, J. & Ziegler, H. Challenges in Visual Data Analysis. Proceedings of Information Visualization (IV), IEEE, p.9-16, 2006.