Beyond 'interestingness' of data

Through a family connection, I ended up having dinner last night with two academics in the field of human-computer interaction. Thankfully, conversation didn't nosedive into the scenario of human-computer marriage - it turns out their study is all about the visualisation of data, and this spurred me into finally getting some of my thoughts together.

There are papers, books, blogs, tweets and more on this stuff, but what's just so enticing for me is the power and ease with which stories can be told when it works. This really is an example of analytic detail and creative visualisation energising and raising each other aloft. Here are some not-that-new examples of visualisations in the digital age.

I'll no doubt come back to this topic but, in the meantime, here's a favourite - and very much analogue - example of Edward Tufte, an inspirational author on the subject of analytical design. Can you spot the seven dimensions depicted in this 1869 map of the Napoleonic retreat from Moscow in the winter of 1812-13?

(For a better image see here)

In no particular order, here are a few thoughts I'd like to explore in this realm (many of which I'm sure have already been extensively looked into just waiting for me to read!)

• To what extent can the growing availability of data and the crowd-sourcing potential of the internet help innovation in this field? Look at the tools already available (sorry for some duplication with the above link)!


• We have all sorts of growing data sets that are based on networks (Facebook, Twitter). How can visualising these and the flow of information across them help us understand the power of communities and the potential social and marketing impacts?


• The widely acclaimed book Nudge is based on a notion of choice architecture and the importance of getting it 'right' for complicated and complex decisions. How can visualisations provide the right feedback and relatability for choosers in certain of these? Maybe an example of this is Fiat's ecoDrive technology.


• In the world of marketing communications, demonstrable effectiveness and efficiency are still the holy grail. Sir Martin Sorrell talks of effective use of insight as one of the critical competencies of his WPP agencies going forward - what can we learn, what can we add, and which start-ups should Sir Martin have his eye on?


• How can this all be used for 'good'? Can the right data and right visualisation of it help to create attitudinal or behaviour change about, say, climate change and energy consumption. Perhaps I've already answered this with the Nudge point above. Alternatively, can we harness or create movements by tapping into the right people? Perhaps the Facebook/Twitter point holds the key to this.

[To explain the title: One common frustration with the way that people look at data is that very often people launch themselves into it in detail without looking at the bigger picture first, missing all sorts of gems along the way. To combat this computer scientists have come up with various measures of 'interestingness', by which a computer can be programmed to pick out different trends, clusters and data points that may prove to be important. But the point is that there has to be a human touch to get at what is actually interesting, and a good visualisation can in some cases tell the critical story with much more power, ease and art than the output of even the most accomplished programme. It has the power to move beyond 'interestingness'...]