Ushahidi: Machine Learning For Human Rights

“2-car acc @ State & Lake, both drivers injred”

That short, hastily typed text message or tweet contains a lot of information that police, emergency responders, news organizations and drivers could use. A human observer could quickly identify that it refers to an auto accident, a medical emergency, and a street intersection in Chicago. But without prior experiences and lots of human input, a computer would likely have a hard time recognizing that State and Lake are streets in Chicago, that “acc” is short for accident, or that “injred” is a typo for “injured.”

The non-profit organization Ushahidi was created by a team of developers so that text messages from eyewitnesses on the ground could be quickly mapped and shared with aid agencies, government authorities and journalists. Since its first use during the violent period after the disputed 2007 Kenyan elections, the organization’s crowd-sourced mapping platform has been shared with human rights volunteers around the world...