GIScience 2014 : [Geographic | Information] Visualization

I gave an invited talk at GIScience 2014 in Vienna where I explored a series of giCentre ideas about Geographic and Information Visualization. I wanted to share a few thoughts and links that relate to the ideas and examples I presented.

Firstly, my original inspiration came from Some Reflections on Cartographic Design in light of digital technology from John Keates (Keates, 1993). I have been conducting rapid "design experiments" ever since.

As I said, I was lucky enough to work with Pete Fisher at Leicester and try to address the views on cartography that he put so eloquently (and strongly) in his Cartographic Journal Paper "Is GIS Hidebound by the Legacy of Cartography" (Fisher, 1993). It was, and still is to an extent in my view. And yet maps seem to work well as flexible interfaces to data that we can change to emphasize the aspects of the data and the phenomena that they represent as we see fit. 

Both of these short papers are well worth digging out.

giCentre visualization work includes a number of methods, models and instantiations that I tried to demo in Vienna. Check bikeGrid to see a grid map (as developed by Eppstein et al., 2012). CSRNet allows you to vary the amount of geography present in a network interactively.

BikeGrid -

BikeGrid -

I showed a concept we call theMapIsTheLegend right at the outset. This was developed during our work on re-thinking legends as described in Dykes et al. (2010).
Our notation for hierarchical visualization - HiVE - and the software that we use to develop HiVE and hierarchical views - HiDE - are documented and available online. They are explained in our 2009 InfoVis paper (Slingsby et al., 2009). We used this combination of method and model to explore bias in the 2010 London local elections in our BallotMaps paper (Wood et al., 2011). 

BallotMap - Wood et al. (2011)

BallotMap - Wood et al. (2011)

Spatially Ordered Tree Maps are explained in Wood & Dykes (2008) and implemented in Jo's TreeMappa library for Processing. Duarte et al. (2004) have come up with what looks to be an improvement on this method at InfoVis 2014 - see their video preview, which explains the method.

OD Maps are introduced in a paper in the Cartographic Journal (Wood et al., 2010).

You can try the PlaceSurvey application online (Slingsby et al., 2014). The same is true of Alex Kachkaev's Survey Glyphs application, which was presented at IEEE EuroVIS this summer (Kachkaev et al., 2014). Alex received the Best Paper award for his work from Kachkaev et al. (2014) from Kachkaev et al. (2014)

The Sketchy Rendering technique is described and evaluated in a paper we wrote with colleagues at INRIA (Wood et al., 2011). Jo's handy library enables you to produce sketchy maps in Processing like this one of achievement by primary school children in London.

Sketchy cartography - produced with the Handy library

Sketchy cartography - produced with the Handy library

Some of the questions that folks asked after the talk were about design process. We try to describe this specifically in most of our papers, but the Creativity paper that Sarah Goodwin led on last year involved deliberate interventions to support creativity (Goodwin, 2011) and shows a model. We ended up with some new metaphors for interaction. Lloyd & Dykes (2011) is quite hard going, but involves a model and recommendations relating to human/user-centred geovisualization design. Watch out for a paper at InfoVis 2014 where we present some broader thoughts on engagement with those who own, use and know data (Wood et al., 2014).

You can also get various utilities, including the classes I used to produce the slideshow from our giCentreUtilities library for Processing

I used the terminology of method, model & instantiation to describe the various artefacts that result from our work rather clunkily as my interest in Design Science is recent and my knowledge  limited - mainly based on Hevner et al. (2004), which started it all of. This is also well worth a read. It provides guidelines for learning through design in a manner that seems to me to have parallels with Mark Gahegan's ideas about abduction and pragmatism. He presented these in a thought provoking GIScience talk on Thursday. 

If you thought the best bit of the talk was Ursus Wehrli's contribution then hunt him down on YouTube. His TEDx talk on Tidying Up Art is hilarious.