Predictive Analytics: Saving Zebras and Other Feats
UK-based conservation charity Marwell Wildlife has been using IBM's SPSS software for predictive analytics to improve its understanding of key issues surrounding the GrÃ©vy's zebra. The endangered species is currently found only in northern Kenya and southern Ethiopia, with only 2,500 of the animals remaining in the wild.
Dr. Guy Parker, head of biodiversity management at Marwell Wildlife, tells CIO Insight that he used IBM's predictive analytics software to study data collected in a recent survey of the local population in northern Kenya. The survey was conducted in collaboration with several other organizations, including the GrÃ©vy's Zebra Trust, the Saint Louis Zoo, and the Denver Zoo. "It's particularly good at teasing out patterns and relationships between different variables - and that's exactly what we wanted to," he says.
For example, Parker says, the software helped determine the factors that drive the local population's attitudes towards wildlife. "We used SPSS to predict what influences people's attitudes, whether that was education level, people's age, what their tribal background is, whether they'd had contact with conservation before, whether they got benefits from wildlife - a whole range of different things," he says.
One of the most important lessons learned from the survey, Parker says, was that zebra hunting by the local population is a key threat. "And one of the reasons for the hunting is people using GrÃ©vy's zebra fats for traditional medicines... so it stands to reason that if we can offer more conventional medicines to people in that situation in the far north of Kenya, that might reduce their need to go out and hunt the GrÃ©vy's zebra," he says. "That could be a fairly straightforward way of overcoming a threat facing the species."
Looking forward, Parker says the organization plans to continue working with IBM to further analyze the data collected. "We'd like to overlay remote sensing data, radio tracking data and aerial survey data, and combine all our information - so we'll be looking to team with IBM to work out how to... feed all these different types of information into our analysis," he says.
And although traditional business measures of ROI don't necessarily apply here, Parker says the results have been exemplary thus far. "We're not dealing in getting a return on our investment in monetary terms - we're dealing in getting it in terms of biological outcomes," he says. "If we can stabilize the GrÃ©vy's zebra populations in northern Kenya, that's our return."