This week I (Guy) have been over in Washington DC at the invite of the Visual Analytics Community to participate in the 2011 Conference. The attendees were mostly US-based researchers and government representatives, plus a handful of industry/commercial types - making for an interesting mix; the US government personnel as the primary 'customer' for much of the work under discussion, with the academics as the primary practioners.

I was invited by the VA team at Middlesex University to join a panel discussion on "Visual Analytics in Practice", alongside US practitioners from industry, government and academia.

My presentation, themed as "tackling the 'too tough' problems in retail and consumer goods" can be seen in the video below:

The panel session itself was very interesting, touching on the requirements of the US National Archives and Record Administration (needing to explore billions of legal documents and emails), analytical learnings at Boeing (including an excellent definition of different types of analytical activity - I hope to cover these in a future blog) and use of Visual Analytics in the retail and consumer packaged goods sectors; despite significantly different domains, key benefits and challenges shone through:

  1. All parties noted huge productivity gains from the use of VA techniques; most describing complex tasks shrinking from weeks to days, or days to hours in some cases;
  2. There was strong agreement on the depth of understanding gained from the use of VA techniques, and that visual exploration left a greater impression on the explorer, and resulted in a more compelling argument presented;
  3. Everyone noted the emergence of a new breed of analytical tools, but concluded that a "one size fits all" solution does not yet exist in this nascent discipline.

The conference itself ran for two days and provided an opportunity for both requests for new tools and technologies, particularly from various sections of US government, and for researchers from universities and national laboratories to highlight new discoveries, software frameworks and analytical techniques. For me, the highlights were:

  • Alan Turner's work on dissecting the visual analysis process, and identify clear disctinctions between analytical activities; leading to better separation of tasks across teams, mentoring of indivudals and formulation of methodology
  • The 'paired analysis' approach of Andrew Wade and Roger Nicholson examining the interaction between subject matter expert and technical expert in joint visual analytics activities
  • David Kasik's classification of 'types' of visual analytics - making a distinction between different purposes, activities and tools/techniques to support operational and strategic analysis
  • Dave Wells closing presentation on the differences between, and emerging role of, visual analytics and traditional business intelligence

Once the relevant presentations become available online I will link them to the points above.

In all, a very useful trip - presentations received, people met and topics covered.  The next phase of the UK VAC kicks off later in May, and Atheon will be attending - subscribe to our RSS feed for news from that event.