Research

IST researchers incorporate social media into emergency response

Tool utilizes Twitter posts to give first responders precise location and vital data

A prototype of the interface that will show geotagged tweets to 911 call center staff to help first responders pinpoint the location of an emergency, discover additional details, and ensure they have the proper resources to save lives in the situation to which they’re responding. Credit: Penn State / Penn StateCreative Commons

UNIVERSITY PARK, Pa. — A team of researchers from the College of Information Sciences and Technology is working to develop a tool that could help 911 call centers integrate crowdsourced social media posts into their operations.

The group is developing and testing software that expands on the information available to first responders upon arriving at an emergency scene. The software will enhance the traditional one-on-one communication between a caller and the 911 call center.

“It’s a new world,” said Andrea Tapia, associate professor in the College of IST and lead researcher on the project, in a recent speech at the iSchool Inclusion Institute. “There is a smartphone in every pocket. Victims and bystanders no longer wait for official communications. And the communications between officials and victims and bystanders are heterogeneous, multichanneled and multidirectional.”

The team plans to optimize social media data to help first responders pinpoint the location of an emergency, discover additional details, and ensure they have the proper resources to save lives in the situation to which they’re responding. 

“When you’re dealing with first responders, it’s critical that they’re as precise as possible,” said Eric Obeysekare, a doctoral student in the college and project team member. “The more information they have before they arrive on the scene, the better they can do their job.”

The researchers recently visited the visited the Charleston County Consolidated Dispatch Center in South Carolina to observe a call center in action and discover how the software could best be incorporated into daily operations. There, they met with call takers, dispatchers, administrators and first responders to understand what information is most critical in an emergency.

“I’d always been under the impression that valid data is most important [in an emergency situation],” said team member Shane Halse, a doctoral student in the College of IST. “But location is. You can have the most valid data in the world, but if you don’t know where it’s coming from, you can’t help them.”

The platform will focus primarily on capturing Twitter data, as a portion of tweets are geotagged — meaning they contain a latitude and longitude to determine the user’s precise location when they sent the message.

“Our software will take all of the tweets from a certain area and show them more specifically on the map,” explained Obeysekare.

He noted that maps are central to 911 call center operations, so it made sense to incorporate tweets into them. Tweets that are not geotagged will be displayed in the sidebar of the software.

“We’re looking for ways to use contextual information from the tweets to have a better idea of what’s happening [in the emergency situation],” he added.

Rob Grace, another doctoral researcher on the team, explained the project’s potential use of social triangulation — a geolocation inference method that attempts to predict the location of networks of social media users in order to collect public data that those users post. This could be useful, for example, if a user posts a message from a location other than where the incident occurred.

“We can filter data by explicit geolocation metadata, by keywords in the content of a post, or we can guess the likely location of a user given the characteristics of his or her social network,” said Grace. “So, if a community would like to collect more geolocated data in the hope of greater situational awareness, social triangulation provides one approach.”

Another benefit of integrating social media into emergency response operations is to give the community another method to request help, aside from the traditional 911 telephone call. Halse noted that the larger the disaster, the more useful the data becomes, and cited Hurricane Harvey as an example.

“911 call centers couldn’t handle the volume, so people turned to social media for help,” he said, noting storm victims who pleaded on Twitter for rescue from their flooded homes.

The team, which includes Tapia, Halse, Obeysekare, Grace and former IST professor Jess Kropczynski, plans to travel back to Charleston in August with a working prototype to test and employ. The group has partnered with Mission Critical Partners, a firm that delivers innovative solutions to enhance and evolve public safety communications, and RapidSOS, an emergency technology company that provides a rich data link from any connected device to 911 and first responders. The team is also partnering with Mines d’Albi, a French engineering school, to implement similar technology at emergency call centers in that country.

Penn State researchers Rob Grace, Andrea Tapia, Shane Halse, Jess Kropczynski and Eric Obeysekare, with Mike Beagles, senior technology specialist at Mission Critical Partners, at the Charleston County Consolidated Dispatch Center in South Carolina. Credit: Penn StateCreative Commons

Last Updated June 27, 2018