The announcement of Jon Gosier’s addition to the Ushahidi Swift River project led to a bit of very interesting speculation from Suw Anderson:
I’m curious to see if there is a reputation system built into it. As they say, this works based on the participation of experts and non-experts. How do you gauge the expertise of a sweeper? And I don’t mean to imply as a journalist that I think that journalists are ‘experts’ by default. For instance, I know a lot about US politics but consider myself a novice when it comes to British politics.
To take a step back, Swift River is a project to “crowdsource the filter” for real-time crisis reporting. Ushahidi provides a platform for aggregating the information around a crisis but, when a crisis situation explodes metaphorically or literally, the information coming in can quickly overwhelm the people trying to make sense of it. Swift River will enable an observer to create a new instance for a given situation, add RSS feeds from various sources including news publications and Twitter, and then additional users will be able to come in as “sweepers” to curate those incoming bits of information and float the most important to the top.
In the comments, Jon mentions that the three “most critical aspects are the trust algorithm (veracity), predictive tagging and filtering out redundancies and inaccuracies.” The first, in my opinion, will be the most challenging, and hopefully most rewarding, piece of the riddle. They’ll be able to scale their ability to float accurate information if they focus on identifying the trustworthy people instead of the trustworthy information.
A couple of weeks ago on Twitter, I observed that the crowd is the least important part of crowdsourcing. More often than not, you could care less about the opinion of the crowd on a whole. What you really want is an authoritative answer, or field report, from the most knowledgeable person in that crowd.