// #Algorithms Use #Twitter and #Flickr to Spot #Flooding - Social Songbird


Latest News


#Algorithms Use #Twitter and #Flickr to Spot #Flooding

Penn State
You would think that satellite images was a precise, effective method of spotting flooding. As it turns out, it's more inexact than you might think. The availability of the images necessary to identify flooded areas aren't always immediately available, and matching them with map data to pinpoint the exact location/see which streets are flooded can be difficult. The solution, it seems, is to add more elements to refine the process.

Twitter has been an instrumental tool during natural disasters in the past and in fact, it was used extensively to help flood victims during the recent monsoons in and around Chennai. That was more about locating people that needed help, and getting the help to them, whereas in this case it's about ascertaining the location and severity of the flooding.

Combining Twitter's data with Flickr, and remote sensors, a group of students from the University of Wisconsin, Penn State and a few others have developed an algorithm-based machine learning system which can find flooded areas with striking accuracy, making the satellite imaging a secondary measure. The system analyses thousands of Tweets and posted images, matching hashtags to map locations. It could also identify water in the images pixel by pixel, in order to figure out which specific parts of the affected location are flooded. Such information could be vital for the people on the ground.

Obviously, it's easy for the human eye to notice where water is present, but for a computer it's much harder, the trade off being that computers can look at thousands of images in a matter of seconds. Other, more specific information in the tweets could also be clustered into a single database, including alternate routes home, changes to traffic conditions and warnings about water levels rising in certain areas.

This is yet another example of social media acting as a fragmented, patchwork data source which when properly filtered can be applied to emergency situations. There is simply too much data for a human team to sift through in reasonable time, but continuing advances in machine-learning mean that it will only become easier, and more vital.

Callum is a film school graduate who is now making a name for himself as a journalist and content writer. His vices include flat whites and 90s hip-hop. Follow him @CallumAtSMF

Contact us on Twitter, on Facebook, or leave your comments below. To find out about social media training or management why not take a look at our website for more info: TheSMFGroup.com
#Algorithms Use #Twitter and #Flickr to Spot #Flooding Reviewed by Unknown on Sunday, January 31, 2016 Rating: 5

No comments:

All Rights Reserved by Social Songbird © 2012 - 2020

Contact Form


Email *

Message *

Powered by Blogger.