Notes from the SXSW I neocartography panel

Andrew Turner, Michal Migurski, Elizabeth Windram and David Heyman speaking at the 2009 SXSW Interactive festival.

Neocartography: Mapping Design and Usability Evolved

Designers are dropping maps into their applications with little concern for usability or design and users are getting "Google Map fatigue." We need to move beyond the simple pin-dropping and consider appropriate mapping interfaces. This panel will look at the current and emerging tools to provide compelling geographic interaction and visualization.

  • Andrew Turner, Mapufacture
  • Michal Migurski, CTO, Stamen Design
  • David Heyman, Axis Maps LLC
  • Elizabeth Windram, Senior User Experience Designer, Google

Please note: These notes were typed and published live. I'll clean them up at some point in the future. Audio forthcoming.

Andrew started with a slideshow of where we’ve come from: past/present comparison. Driving directions from the 20s compared to Google Streetview. Took old concepts and moved to the Web. Where has it worked/not worked. New paradigms for what is possible with cartography. (Will Turner post to SlideShare?)

Red dot fever
When the Google Maps API first really hit the scene lots of web programmers mashed up locations with Google maps and used a red splodge as a location. When you have lots of these symbols on a map you get a screen view that looks as if your PC has come down with measles, hence red dot fever.

Stamen design: A movement down the mapping stack. Looking at new/different ways to render the base map. is great repository for global basemap data. Prior to this, primarily US data. OpenStreetMap is crowdsourced data. People with GPS data and posting government data.

Data quality issues with this—academics and govts do not always approve.

Discussion of choropleth maps.

Cindy Brewer, the accessibility color expert worked on ColorBrewer, color advice for cartography.

Question: Should cartography remain in the purview of experts? There is a large body of knowledge and experience, best practices in cartography. However, no one at Stamen has a cartography degree.

(Definitely need to get more involved with

Stamen switching to this type of tagged, open base map. Outside of the purview of Google, Yahoo! and Microsoft maps.

The EveryBlock project. Hyperlocal information. Decided didn’t need to see street names, because everyone would already know the street names.

80% of males want map oriented with north up, while 80% of females want oriented with direction up (where the direction you’re heading is at the top of the map). [ Should be an option in the user interface; easy to flip ]

What Google Maps has done for usability: click and pan control and slider to zoom are ubiquitous UI elements. And that as zoom in, more information appears or when click on a POI, an info window will appear.

Projections: Some preserve shape, some area, some distance. In WI, each county has own projection for absolute accuracy.

Single-projection world.

Stamen: Translation Buckminster Fuller’s Dymaxion map.

Google Earth, the 3D world. Can be embedded in Web and mobile (iPhone app) now. BBC television centre. Overlay of travel time vs home price. Allow people to figure out where they want to live.

To create really cool maps, also need to start using Flash. To show a non-Flash example, Turner showed us Use open source engine called GraphServer on back-end, not Flash.

Windram: Still thinking about basemap and layer of content over that basemap. Then adjusting the content through various controls.
Stuck in paradigm of basemap v content. Where/how can we move forward, richer interactions between what we consider “content” and “basemap”

Heyman: In terms of richer interactions, use of graphs. Problem with using x/y data for mapping is that can’t use for plotting (scattergraphs). Have the two synced. Filter on graph by attribute then have that change map. Filter on location and have the graph change the map.
Maker statistical package. Correlation analysis. (Find URL)

Migurski: Google focuses on the 80/20 part of the 80, while Stamen focuses on the 20. Google helps people find pizza while Stamen helps explore geo/carto space. This is not a conflict and is good.

Windram: Understanding context and using it to help deliver right information silently and seamlessly. Google trying to move away from just ten results on map when there are a large number of results related to your search/location.

Q: Ability to move/transform data? Interoperability and transformation.
A: Formats like GeoRSS, geoJSON and … Transformation is simple with lat/lon vector data. Raster data is much more complicated.

(Need to finish GeoConvertr)

Q: How to handle when base maps change? The representation of content in time?
A: Stamen starting to address this with some sliders. With UK Olympics. By using temporal slider at bottom, can see content by time. But thinking about expanding to include basemap layer.

(Need to archive data, slices, and then slide over time. Think about with MPROP data)

Andrew mentioned an NYPL project that is based on historical maps/photo data (tweet at Andrew to learn more about this).

Google Earth has ability to slide and move back through time. Oct 1946 to now in San Francisco.

Heyman: Stamen’s London 2012 time slider could become standard UI element for maps.

Migurski: Pirate maps. SF map: only coastline and few other things. Highlevel overview.

Q: Representation of user in crowd sourced map.

Geocommons map of the news

Q: Google Maps sometimes routes you through a high-crime area. Any way to account for crime areas during routing?


(Amazed at the level of design of these maps.)

2 thoughts on “Notes from the SXSW I neocartography panel

  1. Pingback: Posts about Google as of March 16, 2009 | - Make Money Online, Blogging Tips and Reviews

Leave a Reply