Data Lab Link Roundup: African undersea cables, fish volatility, AAAA, outlier detection, cold war maps and 100 interesting data sets

Here are some things that caught our attention last week.

• If you’ve ever seen a map of Africa’s undersea cables, chances are it was made by Steve Song. He just released the July 2015 update of the map and I love this GIF of the map’s history since 2000.  Bonus question: do you know which of the cables was supported by the World Bank? This is literally the “physics of the internet”

• Following the highly anticipated Third Financing For Development conference in Addis last week, we have the Addis Ababa Action Agenda (AAAA)  which contains no fewer than 27 references to “data”. I liked Owen Barder’s take on it - there’s a lot to like in the AAAA, but equally, a lot to do.
  
  
  
  
   

• Netflix briefly broke last week and while they’re best known in data science circles for their 2009 challenge, their techblog is always a great read. They did a fine piece on “outlier detection” and how they automatically tracked down the rogue servers gumming up their systems. I’m now waiting (and available to perform in) a tense action scene in a movie where a data scientist is called in whip together a neural net and save the day.

• Via Twitter, I stumbled across an oldie that I’d totally forgotten about. Robert Jensen’s 2007 paper studying the effect of IT (mobile phones) on the Kerala fishing market. The whole paper’s worth a read but the chart that really strikes shows the effect mobile phones have on the volatility of fish prices:

• Wired ran a beautiful story on “Inside The Secret World of Russia’s Cold War Mapmakers”. Today with products like Open Street Map crowd-sourced street-imagery providers like Mapillary  it’s easy to “forget how hard it once was to come by geospatial knowledge.” And as a bonus, if you’re on a beach in North or South America, this map shows you what’s across the ocean.

• Finally, Robert Seaton offers links to “100 Interesting Data Sets for Statistics” which include everything from 10,000 annotated images of cats, product barcodes and ingredients to global health data, all earthquakes between 1000 and 1903, and 35 million Amazon reviews. I can’t wait to see how these can be combined with the World Bank’s Open Data.