Tag Archives: RSS

Swift River: A tool for organizing Social Media Crisis Information

Post by: Kim Stephens

Although I usually don’t advocate any particular tool,  today I do want to talk about Swift River. Swift enables users to make sense of lots of information across the web in a timely manner. The ability to aggregate data is a key feature that should make this (mostly free) and open-source platform attractive to emergency managers.

Why do emergency managers need to aggregate data from the web and social media sites? The answer is simple, they don’t.  But, if your community is using social media to connect to citizens during non-disaster events, you should expect your citizens will use that same conduit to ask for assistance or to inform the city/county of problems during a crisis. The recent snow storm in the Northeast provided ample examples of how this may occur. Cory Booker, the Mayor of Newark, had an active twitter account before the blizzard–over a million followers; after the snow fell in prodigious amounts, the citizens used this communications platform to let him know about problems they were experiencing.

The pic left has a few examples, most were messages about streets that had not been plowed or problems with signs and fire hydrants. Although it appears he was able to handle the amount of information and questions that he had to sift through during the strom’s aftermath, I think this type of engagement made a lot of emergency managers nervous: How will we be able to keep up with the torrent of requests from citizens during large-scale disasters?

That’s one of the reasons I wanted to introduce SwiftRiver today. In the future–very near future, we will need to be able to do four things quickly:

  1. curate relevant/new situational awareness data as seen from our citizens’ perspective (they are everywhere–our city workers are not)
  2. verify information from non-governmental sources
  3. discard duplicative information
  4. display information in a interactive format with access to and from multiple agencies (including potentially volunteer organizations)

Currently there are not too many platforms that will do all of those tasks, but the folks at SwiftRiver have been working on this concept since March of 2009. From their website:

The SwiftRiver platform offers organizations an easy way to combine natural language/artificial intelligence process, data-mining for SMS and Twitter, and verification algorithms for different sources of information. Swift’s user-friendly dashboard means that users need not be experts in artificial intelligence or algorithms to aggregate and validate information. The intuitive dashboard allows users to easily manage sources of information they wish to triangulate, such as email, Twitter, SMS and RSS feeds from the web.

The key word here in this description is  “users”. Although there is quite a bit of automation in this platform, it still requires actual humans to comb through the data. Some of the tedious work is eliminated, for example the software deletes duplicates or “flags” potential duplicative information, but the act of assigning a veracity score still depends on the user. If your organization does choose to employ a tool like this one, I would recommend that multiple people be trained on how to sort data. (Could this be a job for CERT members?)

This application works well with the mapping tool “Ushahidi” and its developers are members of the Ushahidi team. Ushahidi, at its core, is simply a platform that allows users to place information on a map. You might recall its use during the Haitian earthquake response in which volunteers mapped crowdsourced information regarding damage and injuries.

I expect that in the future there will be lots of competitors for the Swift product, but for now it is difficult to find anything with all of these capabilities. Nonetheless, the website still feels like a start-up.  I’m also not sure how long they will be able to offer the product for free. Currently they still list it as a “free and open source” platform, but their website has a tab for “pricing”.  I’m guessing parts of it will remain free, but organizations, such as  local governments, will probably be expected to pay for the service if they want to take full advantage of all of its capabilities.