Data: The power of Twitter APIs used in disaster detection

The Streaming Twitter API is a read-only stream of status messages that allows developers to detect trends around the web in real-time. The API allows public tweets to be filtered by several factors: user id, keyword, random sampling, geo-location, etc.

As Randall Munroe (creator of xkcd) suggests, the tech savvy users of today are likely to post status messages related to large events as soon as they occur. If the event is large enough, the number of tweets grows, messages get re-tweeted, and #hashtags related to the event are created. This API can be used in many scenarios, but the biggest application is the ability to automate real-time journalism based on topic detection and tracking.

For example, last week's eastern earthquake caused quite a stir across the blogosphere seconds after the tremors were felt. I personally did not realize what I had felt was an earthquake until I began reading my twitter feed minutes after the event. Users from nearby cities in North Carolina to as far as New York were tweeting that they had felt the quake too. Reports of the quake were confirmed once re-tweets of the U.S. Geological Survey's (USGS) message verifying the earthquake's epicenter and magnitude began to appear en-masse.

In turn, the USGS has augmented its earthquake response and alert delivery by creating the USGS Twitter Earthquake Dispatch. The USGS TED gathers information about geo-located reports of shaking using the Twitter Streaming API. This allows the USGS to alert the social web of major earthquakes seconds after they occur as opposed to waiting anywhere from 2 to 20 minutes for scientific alerts from seismic instruments.

Over the years, we have become increasingly connected to social networks and media. Marketing agencies have harnessed the power of easily spreading their products and services these avenues by mimicking word-of-mouth advertising. The Streaming API and USGS TED is just one example of the many ways we can use this method of "word-of-mouth" streams to alert the public of emergencies and disasters.