Spatially Informative Visualization
As sensors, chips and many devices have become cheap to mass-produce over the last decades, we find intelligent systems everywhere. Quite likely, you have a smart phone, which has gone far beyond a device to call with. It is fitted with all kinds of sensors, such as a camera and GPS. It can register data coming from these sensors to form a data. For example, using GPS and its clock, you can track where you have gone on your tourist trip in London! Using this data, you can find out precisely how long you have been in each of the London’s boroughs.
| Borough | Time |
| Barking and Dagenham | 1 |
| Barnet | 3 |
| Bexley | 0 |
| Brent | 1 |
| Bromley | 3 |
| Camden | 5 |
| City of London | 0 |
| Croydon | 2 |
| Borough | Time |
| Ealing | 0 |
| Enfield | 0 |
| Greenwich | 7 |
| Hackney | 6 |
| Hammersmith and Fulham | 0 |
| Haringey | 3 |
| Harrow | 1 |
| Havering | 0 |
| Borough | Time |
| Hillingdon | 0 |
| Hounslow | 0 |
| Islington | 2 |
| Kensington and Chelsea | 8 |
| Kingston upon Thames | 1 |
| Lambeth | 3 |
| Lewisham | 2 |
| Merton | 1 |
| Borough | Time |
| Newham | 4 |
| Redbridge | 2 |
| Richmond upon Thames | 0 |
| Southwark | 6 |
| Sutton | 1 |
| Tower Hamlets | 7 |
| Waltham Forest | 3 |
| Wandsworth | 3 |
| Westminster | 8 |
Now, you still have a day left and you’re planning your tour for tomorrow. You want to go to the boroughs you haven’t been much to yet, but you also don’t want to waste your time sitting in the tube. So, rather than looking at the table, you may want to look at a map instead. In the example below, we colour each borough, such that boroughs we haven’t been to yet, are coloured light blue, whereas boroughs where we have spent more time are coloured dark blue. This is called a choropleth map.
The problem with such a map is that it emphasizes large boroughs, and takes away your attention from the smaller boroughs. So, rather than using the actual borough boundaries, let’s represent each region using a simple square instead. We then stack these squares to make a so-called grid map, ensuring that nearby regions remain close in this very schematic map. And rather than relying on color to encode value (how much time difference is there between two different shades of blue?), we can now very easily encode the data by partially filling the square. We could even show much more complicated data in these simple containers.
This grid map isn’t 100% accurate in terms of geography, but it shows you enough to make an informed decision about a spatial question. We call such visualizations spatially informative. The ALGOVIS project is about algorithms—methods that computers use to compute—to automatically make such spatially informative visualizations.