Using open source mapping tools (Gephi and python’s networkx), I wrote a program that will map the elements of hundreds of stories about any topic you can imagine. It uses data from GlobalGiving’s Storytelling project and a dictionary of all the keywords linked to about 10,000 projects on GlobalGiving’s website to focus the maps on words and phrases that matter to NGO workers.
I think this Information is Beautiful:
2000 stories about “School Fees”
216 stories about “child abuse”
Focusing on what ’causes’ child abuse?
Over 6000 stories from Kibera
A heavily reduced/filtered version of Kibera
Mathare slum in 930 stories
Close up on the core issues for Mathare:
Something more complex? How about 1807 Stories with “hope” in them
A focused view of hope (same map)
Stories are inherently richer and more detailed than surveys and other quantitative data techniques. The problem has been how to visualize the richness while also revealing a specific pattern that can teach us something. This falls short on the prescriptive lessons part, but at least piques my interest to explore all the issues within 40,000 stories from Kenya and Uganda.
Tomorrow I’ll be using these to foster a discussion with local leaders in Mathare about community needs and priorities.