Tweets hold valuable insights amid natural disasters, which the agency aims to leverage.
Social media can be a powerful tool when disaster strikes as personal devices enable people to instantly amplify their experiences to a wider public audience, but according to the National Institute for Standards and Technology, most public safety personnel currently lack the technology and skills needed to effectively harness public, social platforms during emergencies.
In a strategic effort to advance research that it hopes will one day enable public safety-focused entities to tap social media analytics in emergency response, NIST’s Text Retrieval Conference, or TREC, Incident Streams project intends to gain access to Twitter’s Enterprise-Level application programming interface, or API.
“The vast majority of tweets associated with an emergency with a hashtag are wishes of support and discussion about the event,” agency officials wrote in a special notice published Tuesday. “However, on a much smaller scale, people in the emergency itself use Twitter to request aid, share resources, offer services, and provide first-hand reports of the situation.”
Through TREC’s Incident Streams effort, NIST scopes out and prioritizes tweets that occur amid natural crises according to an emergency response ontology, and subsequently invites research teams to develop software applications that can swiftly filter and classify the tweets. The agency produces an infrastructure to gauge how well the created systems perform and ultimately aims to speed up the assets’ improvement. The hope is that once the technologies mature, “such systems could drive analytic dashboards or decision-making processes in public safety organizations.”
However, NIST has used public APIs in this project to date, which insiders note are limited in scope and the data that can be accessed. As a result, the datasets created were ultimately too small to effectively support vast views into relevant social media from a command-and-control center scenario mimicking real-world crises.
“Having access to Twitter's Enterprise APIs would give us the opportunity to dramatically improve the quality of data we can provide,” officials wrote.
The agency aims to examine samples of tweets disseminated by people during emergencies and in-turn use the data to assess novel technologies developed to sift through social media streams as the events take place—and ultimately “route actionable information to public safety personnel.” NIST notes that the types of events it’s looking to collect tweets on and cover are emergencies like floods, typhoons, wildfires, earthquakes, and other natural disasters. Officials also specify that happenings that arise out of public controversy, protests, riots, strikes and demonstrations will not be captured in the effort.
Twitter “does not provide direct access to the federal government via contract,” NIST said, but the notice of intent confirms that the tech company has permitted the New Jersey Institute of Technology or NJIT of Newark, New Jersey to act as an intermediary for the effort.
“Twitter is unusual among social media companies for providing this kind of API data for researchers and companies,” NIST noted. “Neither Facebook, Instagram, nor WhatsApp provide a comparable API Data.”Though the agency clarifies that the special notice is “not a request for competitive quotations,” responses to the notice of intent will be accepted until July 16.