Continuing with examples of summarizing information using the wordtree algorithm, here is what 768 HIV/AIDS projects on GlobalGiving talk about (red dots) contrasted with 4,231 stories that mention HIV or AIDS (blue dots). Words that both organizations and stories use are shares of purple/pink:
Interesting overlapping words [alignment between HIV stories and HIV projects]: parents, girl, boys, hospital, women, patients, medical treatment, services, children, rights, providing, living, hope.
Projects focus more on: food, youth sports, child, vulnerable, skills, clinic, Kibera, local, social, water, primary and secondary schools, thank.
Communities focus more on: counseling, drugs, sex, condoms, abstain, science, behavior, testing, income generating activities.
Here is slightly larger summary map. Same coloring scheme, but with the expanded version, individual organizations start to branch out and diverge into their unique perspectives on HIV/AIDS:
This last image displays some distinct branches related to HIV/AIDS projects on GlobalGiving. VAP, Mfangano, CFK, Sadili, Jackson, Freeplay, Kutamba, St. Vincents, and WMI are all organizations or project names. The pink words form a list of words and concepts that organizations should be aware of when designing and managing projects that target HIV in a community. While this doesn’t directly identify root causes to HIV related problems, as this other approach would do for HIV, it is a first step.
Update: BigML analysis of 4,217 HIV/AIDS stories
BigML allows one to see which of the many survey characteristics associated with each story are most important (in the sense that they “explain” variability in the outcomes of stories. Possible outcomes are ‘success’, ‘failure’, or ‘mixed’.
HIV Success story pattern #1: Storyteller felt happy about story, were between age 31 and 45, did not talk about “fun”, and were from Uganda, not Kenya.
HIV Success pattern #2: Felt hopeful, Uganda, story less than 491 characters in length, and events happened 1-2 years earlier.
HIV stories with mixed outcomes: Inspiring stories from observers about problems.