Mapping Post-Election Violence in Kenya

Posted by Brannon Cullum in Africa, Citizen Media, Partly Free, Poorly Wired, Location Based
originally posted on June 24, 2010

THE CHALLENGE

In early 2008, rioting and looting erupted across Kenya, as many Kenyans believed that recent presidential elections were rigged. There was much confusion and insecurity due to a government-imposed media blackout. How could Kenyan citizens learn about incidents of violence and become more aware of what was happening on the ground?

 
The Players
Ory Okolloh, a native Kenyan who blogs [http://www.kenyanpundit.com] about her country’s politics, wrote about what was happening and asked readers to email her or post comments with details about incidents they witnessed. She soon found herself overwhelmed with emails coming in from a variety of sources that were sharing updates about the situation. On her blog, Okolloh asked if there were any other tech-savvy people out there interested in creating a mash-up showing where incidents were occurring on a map. Bloggers and developers, including Eric Hersman (White African) [http://www.whiteafrican.com], David Kobia (DKfactor) [http://www.dkfactor.com], and Juliana Rotich (Afromusing) [http://afromusing.com/], quickly came together to create a new resource for all Kenyans. Co-founder Eric Hersman found himself asking: [http://www.africanews.com/site/list_messages/14259%3E]
“How about a platform that serves as a centralized repository for on the ground reports from any Kenyan via SMS? The ability for people to upload videos and images with some text to a web-based and mobile phone accessible site?”
Ushahidi [http://www.ushahidi.com/] (which means “witness” in Swahili) was born.
 
The Tools and Tactics
The Ushahidi platform was developed as a rapid prototype model that enabled individuals, as well as members of NGOs, to submit reports via SMS or email detailing acts of violence and trouble spots. A Kenyan could send an incident report with location details to a short code number. The text or email would be rerouted through FrontlineSMS [http://www.frontlinesms.com/] and synched with the Ushahidi platform. Then, the message would be received by an administrator who would attempt to verify the information with the original sender. If the report proved credible, it would be uploaded onto Google Maps in as close to real-time as possible. What resulted was a map [http://kenya.ushahidi.com/main] populated with aggregated reports of incidents of violence and looting and the identification of places in need of aid relief.
The platform helped to fill a void left by the mainstream media and provided Kenyans with information vital to their safety and peace of mind. The strength of Ushahidi was found in its users’ ability to crowdsource information onto a Web-based map. Eyewitnesses were no longer limited to being bystanders as the crisis unfolded, but were able to become actively involved in reporting incidents.
 
The Stumbling Blocks
One of the primary challenges encountered during the first deployment of Ushahidi was verifying and authenticating reports. How could administrators identify reliable information? The Ushahidi staff did not have a formal verification process at the time; members would call or email reporters to verify their accounts and try to counter-check stories with other sources or reports coming in. If they were unsure about a particular report, it would be noted on the website that the story had not been verified. Also, each report had to be uploaded manually onto the Web-based map by Ushahidi staff, which proved to be time consuming.
 
The Outcome
The use of Ushahidi in Kenya was successful in drawing increased local and global attention to the crisis, as well as in filling the gap left by news media organizations that had been silenced. The platform had 45,000 users in Kenya during this time of turbulence. Radio deejays read some of the reports on air.
A study done by the Harvard Humanitarian Initiative [http://irevolution.wordpress.com/2008/10/23/mapping-kenyas-election-violence/] analyzed data reported on the site and compared it to reports from mainstream media outlets. The study concluded that Ushahidi was better at reporting incidents as the started (rather than just the deaths resulting from incidents) and reports covered a broader geographical area than those coming from mainstream media.
Since its first deployment in Kenya, Ushahidi has become a nonprofit company and its members have built a platform [http://download.ushahidi.com/] that is offered for free to independent organizations. The open source software can be modified and adapted for various situations and contexts. It has been used around the world, including in South Africa (anti-immigration violence), eastern Congo (violence), Mexico (election monitoring), India (election monitoring), Gaza War (crisis monitoring), Haiti (post-earthquake), Washington, D.C. (Winter 2010 snowstorms), and Chile (post-earthquake).
In efforts to combat false reports, Ushahidi now offers a brief verification guide [http://www.ushahidi.com/media/c_Ushahidi-Verification-Guide.pdf] for its community members. Patrick Meier has written extensively on his blog [http://irevolution.wordpress.com/2010/04/08/wag-the-dog/)] about the issue of falsifying crowdsourced data, and he spearheaded the development of another software, Swift River [http://swift.ushahidi.com/], that uses crowdsourcing and algorithms to validate and filter news.
What started as a project built from the ground up has become a scalable and sustainable nonprofit technology company due to a model platform that can be modified to suit end users’ needs. Where will Ushahidi be used next?
 
Key Takeaways
1)    Seek out individuals who are willing to collaborate with you and offer their expertise.
2)    Keep tools simple. You often don’t need expensive software or devices to achieve your goal.
3)    Don’t underestimate the willingness of citizens to be engaged and participate.
 

 

 

The Challenge

THE PLAYERS

Ory Okolloh, a native Kenyan who blogs about her country’s politics, wrote about what was happening and asked readers to e-mail her or post comments with details about incidents they witnessed. She soon found herself overwhelmed with e-mails coming in from a variety of sources that were sharing updates about the situation.

On her blog, Okolloh asked if there were any other tech-savvy people out there interested in creating a mash-up showing where incidents were occurring on a map. Bloggers and developers, including Eric Hersman, David Kobia, and Juliana Rotich, quickly came together to create a new resource for all Kenyans. Co-founder Eric Hersman found himself asking, “How about a platform that serves as a centralized repository for on-the-ground reports from any Kenyan via SMS? The ability for people to upload videos and images with some text to a web-based and mobile phone accessible site?”

Ushahidi—which means “witness” in Swahili—was born.

THE TOOLS AND TACTICS

The Ushahidi platform was developed as a "rapid prototype" model that enabled individuals, as well as members of NGOs, to submit reports via SMS or e-mail detailing acts of violence and trouble spots. A Kenyan could send an incident report with location details to a short code number. The text or e-mail could be rerouted through FrontlineSMS and synched with the Ushahidi platform. Then the message would be received by an administrator who would attempt to verify the information with the original sender. If the report proved credible, it would be uploaded onto Google Maps in as close to real time as possible.

What resulted was a map populated with aggregated reports of incidents of violence and looting and the identification of places in need of aid relief.

The platform helped to fill a void left by the mainstream media and provided Kenyans with information vital to their safety and peace of mind. Eyewitnesses were no longer limited to being bystanders as the crisis unfolded but were able to become actively involved in reporting incidents.

THE STUMBLING BLOCKS

One of the primary challenges encountered during the first deployment of Ushahidi was verifying and authenticating reports. How could administrators identify reliable information? The Ushahidi staff did not have a formal verification process at the time; members would call or e-mail reporters to verify their accounts and try to counter-check stories with other sources or reports coming in. If they were unsure about a particular report, it would be noted on the website that the story had not been verified. Also, each report had to be uploaded manually onto the web-based map by Ushahidi staff, which proved to be time consuming. The Ushahidi team has since addressed these problems, although each new use of the tool faces its own unique set of challenges requiring its own set of solutions.

THE OUTCOME

The use of Ushahidi in Kenya was successful in drawing increased local and global attention to the crisis, as well as in filling the gap left by news media organizations that had been silenced.

The platform had 45,000 users in Kenya during this time of turbulence. Radio deejays read some of the reports on air.

A study done by the Harvard Humanitarian Initiative analyzed data reported on the site and compared it to reports from mainstream media outlets. The study concluded that Ushahidi was better at reporting incidents as they started (rather than just the deaths resulting from incidents) and reports covered a broader geographical area than those coming from mainstream media.

Since its first deployment in Kenya, Ushahidi has become a nonprofit company and its members have built a platform that is offered for free to independent organizations. The open-source software can be modified and adapted for various situations and contexts. It has been used around the world, including in South Africa (anti-immigration violence), eastern Congo (violence), Mexico (election monitoring), India (election monitoring), Gaza War (crisis monitoring), Haiti (post-earthquake), Washington, D.C. (winter 2010 snowstorms), and Chile (post-earthquake).

In efforts to combat false reports, Ushahidi now offers a brief verification guide for its community members. Patrick Meier has written extensively on his blog about the issue of falsifying crowdsourced data, and he spearheaded the development of another software, SwiftRiver, that uses crowdsourcing and algorithms to validate and filter news.

What started as a project built from the ground up has become a scalable and sustainable nonprofit technology company due to a model platform that can be modified to suit end users’ needs. Where will Ushahidi be used next?

blog comments powered by Disqus