Julia Evans

Open Data Exchange 2013

I went to an amazing conference yesterday called Open Data Exchange.

There were so many interesting, thoughtful people who are really interested in how to make the best use of open data to make things better. It was the best open data event I’ve ever been to. Here’s a few things that stood out to me. I’ve probably misrepresented what all of these people were saying in one way or another, but hopefully it is more or less accurate.

One point that came up again and again is that it’s important to humanize data and to connect programmers and designers and open data advocates with citizens, business owners, and experts in their fields. Open data apps are only useful if they’re solving problems that people are really having, and answering questions that people really need to know the answers to.

Lightning talks

  • Diane Mercier from the city talked about her work in educating people who work for the city on how to open up data. She’s amazing and I’m so glad that there are people like her advocating for open data at the city.
  • Trina Chiasson talked about how important it is to humanize data, using the example of the piles of documents collected during the Guatemalan civil war. This really made me think about how to make sure that things I work on are actually useful to people, and not just fun to work on.
  • Anton Dubrau demoed an app he made that overlays 1947 aerial photos with 2011 satellite views. His was the only talk to get spontaneous applause. It’s definitely worth checking out his blog on maps and public transit.
  • Alex Aylett from Eco Hack MTL talked about urban sustainability projects. He’s interested in how open data can apply to organizations like Santropol Roulant and Alternatives. He said that his main goal with Eco Hack MTL was not to work on any specific project, but to connect people from different areas and get them to talk to each other. Fantastic. There’s going to be a 5 à 7 in May and I’m definitely going to be there.
  • Roberto Rocha, a data journalist at the Gazette, talked about how programmers can collaborate with journalists. In particular, he said that journalists
    • can help ask the right questions to put together a story
    • have an audience :)
    • know more about how to talk to politicians and organizations to get information, including how to make access to information requests
  • Edward Ocampo-Gooding from Open Data Ottawa introduced the idea of speed idea dating. I’m definitely going to try this when I next organize a hackathon-type event. They also organized a meeting where programmers got to talk to the local politicians who were providing the open data. So impressed with what they’re doing and how open the local government in Ottawa is to these initiatives.


Pete Forde mentioned this blog post by David Eaves, about how open data saved Canada billions of dollars.

People often talk about the potential open data has to uncover corruption, and it’s really great to see concrete examples of that.

Josée Plamondon who works on ContratsNet spoke about how she gets citizens, business owners, and experts involved. She said she had a really valuable conversation about the issues with contracts in Montréal with someone who owns a small construction company who explained what they find frustrating. She’s found that it’s really difficult to engage people online, and if you have meetings in real life you can learn a lot more. Technology is less important than asking the right questions

Corey Chivers talked about work he’s doing to find out how the results people are reporting in ecology papers are changing over the years. (slides)

One of my absolute favorite things was Tracey Lauriault’s talk on “Geospatial Data Infrastructures” – how geospatial data is being organized, and how we can learn from that. A couple of notes I took during her talk:

  • Groups like the Group on Earth Observations have huge amounts of data and have been thinking about how to manage it for a long time. We can learn from them.
  • Geographers have always had to share data, because mountains don’t respect country boundaries. They’ve been doing this for decades.
  • policy organizations don’t always look at data analysis, data analysis people don’t always think about policy

Transit maps: 1941 vs 2011 PyData Boston 2013