Live-Blogging Tech@State: Ignite Sessions

Clay Johnson, author of “Information Diet.”

The U.S. started with a constituent to representative ratio of 60,000:1 and is now somewhere over 700,000:1. There will be a crisis if government can’t take the huge and growing amount of information from people and interpret it and use it in an effective way. With more and more channels popping up, the volume keeps getting turned up. When the volume is up too high, it’s impossible to hear, making the government deaf. New tools are coming into being to help channel the voices into government, but what we need are ways to help them hear better, not help us project louder. Instead of building better megaphones, let’s build better headphones.

Adam Beaugh – PolitEar

If you said 1 year ago that there would be an upcoming (June 2012) Egyptian presidential election with 41 parties and 11 candidates, very few might believe you. What’s more, is those candidates are interacting with the Egyptian public through social media, marking a shift from social media as a tool in revolutionary action, to a tool for conducting democratic business. There are a lot of moving parts and lot of real-time information flowing, which can lead to information overload and a need to digest and filter. Democracy can be very messy and unorganized. Sometimes complexity needs complexity to create simplicity.

Brian Keller – SAIC

Biometrics are the measurement of features or behavior that can identify a person (iris, fingerprint, heartbeat) and have a host of applications in the government. In 1805 Army was first service to use bio-metrics (to identify soldiers). First responders could use to biometrics to identify who is on the scene and whether they should be there. Border patrol can use hand held devices to determine who is crossing the border. USCG is using handheld biometric readers effectively to determine who is on ships, whether they be cargo crews, drug smugglers, or immigrants.

What’s the case for using biometrics? Strengthened security and, you could argue, time and money savings. However there are always civil liberties issues that must be considered whenever implementing biometrics.

Robert Kirkpatrick – UN Global Pulse

Kirkpatrick built on his earlier discussion do explore how you can use sentiment analysis for predictive purposes. In Ireland, spikes in confusion preceded spikes in anxiety. In US, there are spikes in anger, followed by discussion of people needing to downgrade housing or move to public transportation. In Indonesia, spikes in discussion of food track with Ramadan. Elsewhere, tweets about the price of rice formed a curve that nearly mirrored the government price indexes.

Katie Baucom – National Geospatial-Intelligence Agency

Social media has been a valuable tool for more information for an agency that has traditionally dealt with satellite imagery. Pictures of damage on Flickr, videos on YouTube, reports on twitter can all be used to help direct resources. With social media, you can correlate density of population with density of activity on social media and calls for assistance, to help prioritize response. Additionally, it’s important to try to get data out to responders, too, not just pull it in. Mobile apps can help responders access data that would be available to them in a field office. The challenge for a traditional imagery agency is to make a move towards trusting public data and reports.

Will Law – Netvibes

Law showcased his dashboard, which was created to help private sector organizations monitor brands and products. He says while there is a lot of hype around the “fire-hose” or data, but that it’s really important to step back and personalize the content that is relevant to the user and what she is trying to achieve. Since users can often do this better than an algorithm, users take the first stab at selecting what’s important to them. Once those preferences are defined, you perform analysis, based on the what, how, who, and where. By allowing users to manipulate data and see sentiment, by volume, over time, you can start bringing out some interesting findings. When applied to Syria today, Law had a couple findings:

  • By far the biggest volume of social media content was on Twitter
  • Using word cloud, it was very easy to identify the cities where fighting was most intense

Law closed with this (often over-looked) note: “Real-time is only useful if someone is there to see the data changing.”

Leave a Comment

Leave a comment

Leave a Reply