‘Network Visualization in an Age of Interconnectedness’ – Manuel Lima Talks at BBH

A talk on Data Visualization by Manuel Lima, courtesy of @madebymany and @bbhlondon, 25th August 2009.

“I am a Functionalist Troubled by Aesthetics” – Manuel Lima, quoting Wim Crouwel

Ostensibly Manuel talks about representing data, visually: at a more abstract level, he is talking about the transmission of information across a continuum from raw data, to a global, understandable ‘information’, to a communicable ‘knowledge’, and finally to a personal ‘wisdom’. This process remains contingent on context, and on the relationship between production and consumption of data.

Firstly, Manuel spoke of the ways in which information has been communicated throughout the ages – considering cave drawings, moveable type, and of course, language. Contextualising his topic, he moved on to what he referred to as a ‘Visualization Outburst’ – brought about by five key factors.

1. The capacity of digital to store information.
References to the exponentiality of digital growth and ‘Kryder’s Law‘ were made, examples being the capacity of the iPod to store 160GB of information in 2009 compared to the benchmark model in 2001 boasting 5GB capacity. Kryder’s Rule states that the capacity of information storage will double every 18 months, and this has been proven in several cases: the iPod being one, and to name another, Manuel suggests that a laptop computer will have a commodity drive capacity of 1 Petabyte (1 million Gigabytes) by 2030.

2. Openness of Datasets
In sharing data (eg IBM’s Many Eyes), and allowing others to access your data, as well as being able to aggregate multiple users data (eg through APIs), we have more data to work with: and more is better – although perhaps more complex. An onus on transparency and openness is championed in many contexts, and the manipulation of data by third parties can be mutually beneficial – as well as forward facing.

3. Social Networks

The interconnectedness of individuals in an online capacity has a huge impact on information sharing. Not only who is connected to who by eg a LinkedIn profile, six degrees of separation and so on: but also through topics of interest, communal activity, music tastes and so on. Further, the aggregation of user data by host platforms such as Twitter, and the APIs they provide, are sources for data viz in their own right. Tag clouds used within the flickr platform were early and benchmark examples of democratized data visualization.

4. Democratization of Tools

Further exploring the democratization of data, we are shown examples of software such as processing and flash facilitating UG data viz. Data visualization is no longer confined to academic field, but can become part of a wider conversation of users and resultantly is allowed to form the syntax/discourse/language for communicating data across disparate platforms.

5. Mainstream Media
Along a similar line, Manuel talks about ‘vernacular visualizations’, and a point is raised from the audience about how whilst openness of data-sets drives data visualization, similarly the dissemination and adoption of data visualization drives the opening of data-sets. Manuel rebuts along the line that sharing information is about an exploration; a journey – a return to the link between producer and consumer – and that the objective is to provide a greater explanation; a function of the data, rather than visualization being an end in itself. I shall return to this point later.

Visual Complexity: We need to make a transition from Tools of Curiosity to Tools of Functionality”

California NanoSystems Institute

California NanoSystems Institute

Secondly, Lima moved on to discuss his own project, Visual Complexity. Here he moved through several fascinating examples of how data viz is being used, with an emphasis on plurality: in working towards a ‘common language’. His work and curation encompasses a variety of fields, where the data he has collected might be from biology, social networks, business, IT, music, politics or astrophysics. The point in many ways is that the subject matter, doesn’t matter: the interpretation of the viz is subjective and entirely within the control of the user. The job of data viz is to make that interpretation clearer or more valuable than through other methods.

Within this section he gave several examples, the highlights of which for me were a project on GPS drawing, whereby children would walk around a large open space and physically ‘draw’ a simple object, the example being an elephant’s head. They could then map this data using GPS technology onto the terrain which they had navigated, and remove sufficient data to create a real-life, mass-participant ‘art attack’. Neil Buchanan would have been proud. Other examples included linking last.fm music tastes across a social network and so on, before Manuel moved on to more hardcore applications of this hybrid of design and technology (‘a new science’). He alludes to several examples of visual representation of terrorist networks, identifying key players over a temporal space, as well as analysis of the demise of Enron, a stab at the ‘fat cats’ of the US and many more – all of which are available on his blog.

“Aesthetics should be a Consequence, not a Goal of Data Visualization”…

Tracing the Visitors Eye

Tracing the Visitor's Eye

My personal favourite combined UGC in the form of geotagged flickr photographs, geographic data and temporal data to create a map of the paths people take around a space: in this case, the city of Barcelona. Basically, the data aggregated user photos of landmarks in the city, and pitched them against the time they were taken, to establish the routes people created: “Tracing the Visitor’s Eye”. This data was then overlaid on a map of Barcelona to show the traces people had taken, with stunning informational and aesthetic effect.

“Time is a Very Difficult thing to Map”

The final part of the talk focused on network visualization, and how this translates into every day life. Examples were given of visual representation methods, as well as interactive exploration techniques, ranging from radial convergence models and radial centralized networks, through to multi-sensory installations such as a Californian project involving the representation of nodes in both colour and sound, across a three storey building.

So. What has all this got to do with me? Well, to be honest, I wasn’t 100% sure, but all I knew was: it looked fucking cool, and seemed to serve a real functional, valuable purpose in today’s data rich and time poor society. However, I had an inkling that my work looking at creativity within systems, and as processes, would relate – especially in terms of how best to represent this visually. In his final section, Manuel alluded to the complexity of networks, considered from different levels of aggregation, and how data viz can serve to simplify models of complex phenomena.

He splits the system up into three levels: the macro level, or ‘system level’, where data serves to indicate patterns; a relationship level, where connectivity between nodes is of focus, and the micro level, looking at individual nodes. I map this on to my own work, discovering parallels within a creative system of Culture as System (Macro); Organizational (Relationship) and Individual (Node) level aggregation.

Further, his ideas on time, and how this is difficult to map were very interesting. I attempt to map the development of creative projects over time both through language; and diagrammatically. Lima uses an example of a temporal based network visualization, showing schoolchildren interacting with a teacher, and one another, over time. It blows my mind, despite its relative aesthetic modesty. But his point here is that nodes within this system can and should be represented intelligently: incorporating temporal factors, factors of relevance (eg proximity to other nodes: there’s no need to see the whole system, actually, only what’s relevant), and factors of simplicity.

He closes with an allusion to a ‘Universal structure’ – which he translates as matching images of neurons in a mouse’s brain to the Millennium Structure, developed recently to represent the universe-as accurately as we can, using a self contained data set of 25TB.

He jokes at the juxtaposition, but similarity, between one of the smallest things you can see, and one of the biggest things you can think of: but to me, and probably to many who were there, the message was clearer: working towards data visualizations or representations of the monolith of data we face is not an end in itself, but a stepping stone in creating a multi-disciplinary and cross discourse platform for communicating information: be they ideas, data-sets, evolutionary systems or, indeed, Facebook friends.

Manuel Lima | Visual Complexity from digup.tv on Vimeo.

Anyone who was at the event, please contribute to my thoughts. Manuel: keep up the good work.


Define the Space: It’s all about Context #2: Interpreting Creativity through Changing Epistemologies

Duchamp's Fountain

Aesthetics, for instance, has traditionally been the study of how to tell art from non-art and, especially, how to tell great art from ordinary art. Its thrust is negative, concerned primarily with catching undeserving candidates for the honorific title of art and keeping such pretenders out. The sociology of art, the empirical descendant of aesthetics, gives up trying to decide what should and shouldn’t be allowed to be called art, and instead describes what gets done under that name. Part of its enterprise is exactly to see how that honorific title–“art”–is fought over, what actions it justifies, and what users of it can get away with. (See Becker 1982, pp. 131-64.)

From ‘The Epistemology of Qualitative Research‘; Howard S. Becker