Baltimore residents know they live in a racially segregated city but, thanks to a map of Twitter activity developed by local tech entrepreneur Dave Troy, they can now see just how divided the city is socially as well.

Troy grouped people in Baltimore, and other cities around the world, by whom they follow and talk to on Twitter, then worked with local collaborators to analyze what those people talk about most. He looked at who users’ friends and followers are and what they Tweet about.

The resulting map, in the case of Baltimore, shows a city with two extremely separate communities, white on one side, black and Latino on the other, but also with starkly different online areas of interest and little socially in common besides sports.

Locals refer to the town as “Smalltimore,” Troy explains in the TED talk video, featured on their site today “because it seems as though we’re living in a very small town. We see the same people over and over again.”

“But that’s because we’re not really exploring the full depth and breadth of the city,” he observed. Other cities map out with much more heterogeneous populations.

Subgroups in Rio de Janiero, for instance, have lots of overlap, coming together in particular about different kinds of music.

Troy says he hopes the people-mapping concept, and Baltimore’s profile in particular, can stimulate a helpful conversation that breaks down social barriers: “Only looking at race doesn’t really contribute to our development of diversity.”

Describing a city as “the sum of the relationships of the people that live there,” he said, “if we can start to document those relationships in a real way then maybe we have a real shot at creating the kind of cities that we’d like to be.”

Troy’s Take-aways

Asked today what the key discussion points for Baltimore might be, Troy sent these thoughts:

1) The social network map roughly mirrors the population makeup of the city, with a very large percentage of African Americans (and other minorities) represented. So it’s not badly biased against minorities, which is one suspicion people have about network data.

2) In fact, minority usage of Twitter is higher than for “white” populations (some data here).

3) My findings globally suggest that social network use is inversely correlated with income; the more secure they are the less, they feel the need to network. This makes sense and the findings bear this out in everyplace I’ve studied.

4) The eccentricity of the network (compared to say, Barcelona) shows we have a very divided city. If we measure this eccentricity over time, we can get a feel for how we might be doing at mending that gap. We have a lot of work to do.

He also noted that Baltimore is not the only American city with one of these big, wide profiles of non-overlapping-ness. He’s about to unveil another city map that’s been in the news lately: Greater St. Louis, including Ferguson.

“It looks a great deal like Baltimore,” Troy said, “and that should give us some cause for concern.”