Crowdsensing: mobile data and predictive algorithms

In Pakistan, mobile data has helped the authorities predict where an epidemic will break out:

Researchers working for the Pakistani government developed an early epidemic detection system for their region that looked for telltale signs of a serious outbreak in data gathered by government employees searching for dengue larvae and confirmed cases reported from hospitals. If the system’s algorithms spotted an impending outbreak, government employees would then go to the region to clear mosquito breeding grounds and kill larvae. “Getting early epidemic predictions this year helped us to identify outbreaks early,” says Umar Saif, a computer scientist at the Lahore University of Management Sciences, and a recipient of MIT Technology Review’s Innovators Under 35 award in 2011.

When we think about “mobility” and its potential in business and society, we shouldn’t limit ourselves to the desktop and app paradigm. (more…)

Ushahidi: realtime social media lessons from crises (and a model for slow news?)

Yesterday, at the Legatum Institiute‘s Next Generation Philanthropy Forum I got to meet Juliana Rotich, programme director at one of the most interesting open source projects in the world, Ushahidi.

If you don’t know it, Ushahidi is an open source platform for communicating in a crisis. At simplest, it is a way of aggregating text messages, emails, Tweets, blog posts and mainstream media articles to form a clearer picture of what is happening in a fast moving situation, say in a war zone or a natural disaster. It’s also been put to good use in places like the Lebanon and Mexico by people wanting to monitor the fairness or otherwise of their own elections and to help with the effective distribution of vital medicines in Malawi, Kenya, Zambia and Uganda.

So on a geeky level Ushahidi’s obviously fascinating, on a humanitarian level it’s seriously inspiring, but there are lots of other elements of the project which are useful to consider.

Listening to Juliana’s presentation (I understand the videos will be live on the Legatum site soon) and chatting to her in the break, there a few notes I made that I will share here:

  • Web as witness: This is my take on what Juliana was saying, but I got a sense that by managing information on Usahidi served both as resource for people involved in it but also to put things on the record. As Juliana put it: “If a tree falls in a forest and Google doesn’t hear it, does it make a sound?”. Ushahidi means “testimony” in Swahili – so I guess this purpose has been baked in to the development of the platform.
  • Spreading social web beyond developed countries: Juliana is interested in how you stop “social media becoming an enclave for developed countries”. There are so many talented developers and creative people in the developing world, and she wants “to invest in those minds”. You can see her point – a massive latent cognitive surplus, to borrow Shirky’s phrase, is in developing countries, with all its incredible potential denied to the networks for now. If I was a VC with a long view, I’d think about heading for Africa…
  • Mobile is key to connecting the developing world: This is not news, I appreciate, but mobile handsets and access are the way that the developing world can connect right now. As Jared Cohen of the US State Department said in his speech earlier in the day, the economy of Kenya is so reliant on mobile payments that it would collapse tomorrow if you were to remove the GSM network. The mobile is the “default device” for Ushahidi’s developers, said Juliana. She also lamented that Twitter lacks a text message interface [I paraphrase]: “With SMS Twitter could become the pulse of the whole world – not just the developed world.” Now there’s a thought…
  • Near-realtime filtering: A major challenge for Ushahidi is filtering information as it comes in in near realtime. There may be disinformation from antganists, but also incorrect information, and alot of echo (re-tweets count as this) and maybe spam. Ushahidi uses manual filtering, Akismet and Swift River, a kind of crowd-filtering approach which “rescues data from the river and puts it on the bank”. This process involves a lot of human intervention at the moment, but they are working on algorithms to automate a lot of this.
  • Realtime media with a slow news legacy?: It strikes me that the combination of fact-checking and contextualising of realtime information is an immediate benefit of Ushahidi, with emergent benefits being that complex data has been curated which can be used by journalists, NGOs and others who want to analyse and learn from a crisis later on. This model is maybe how news organisations need to think about their role. I first heard about Ushahidi via the Al-Jazeera Labs project using the platform during the recent war in Gaza. Is there a case for the BBC to run a similar model when breaking news hits, or for news organisations to cooperate with a Usahidi like model to make sense out their reports and the mix of witness accounts on the ground?

There’s more on this approach in this video, which highlights the danger of rumours in a situation like the Mumbai terrorist attacks:

An Introduction to Swift River from WhiteAfrican on Vimeo.


Swift River looks like it could be a very important development, not just for Ushahidi, for everyone living with the explosion of data brought about by the realtime web. There are obviously lessons here for news organisations and others (i.e. most organisations and communities of interest).

: : Bonus link: I’ve posted some more notes and my presentation slides from the Next Generation Philanthropy Forum on the iCrossing UK blog.

: : One more thought. Friends of mine in NGOs have told me before and the theme came up again yesterday that it is impossible to micropayments efficiently online because of the cost of transactions. That is to say, if I donate £1 to UNICEF online at least 21p of that pound will be lost to the transaction cost in the *best case*. PayPal or Google Checkout should develop a charity / NGO model – imagine how much money could be freed up for NGOs they were able to ask millions of people to send a few pence or cents?


How we blew up the economy

Thinking about banking, finance, the web and innovation in general at the moment and came across this rather fine explanatioin of the credit crisis from Jonathan Jarvis, an artist. If you’ve got ten minutes and want for once and for all to understand how we got into this mess, then have a watch…


The Crisis of Credit Visualized from Jonathan Jarvis on Vimeo.
Via Free Exchange