UNIVERSITY PARK, Pa. — A group of faculty and graduate students from Penn State's College of Information Sciences and Technology recently shared their research at the International Conference on Information Systems for Crisis Response and Management (ISCRAM). The conference was held May 20-23 in Rochester, New York.
Research presented at the conference included contributions from Andrea Tapia, associate professor of IST; Jake Graham, professor of practice in IST; Guoray Cai, associate professor of IST; visiting scholar Lida Huang; and doctoral candidates Bill Aurite, Jomara Binda, Rob Grace, Shane Halse, Nick Lalone, Feng Sun and Samantha Weirman. Jess Kropczynski, an assistant professor at the University of Cincinnati who formerly served as a postdoctoral researcher with the college, accompanied them.
ISCRAM is an international community that promotes research and development, exchange of knowledge, and deployment of information systems in the field of crisis management. Now in its 15th year, the conference explores research related to crisis informatics, geographic information systems, humanitarian information management, and situation awareness.
Throughout the conference, IST researchers shared their findings and perspective through eight papers, a poster presentation and a keynote address. Cai and Sun, along with four other co-authors, were awarded the conference’s best research paper for their work, “Modeling Threats of Mass Incidents Using Scenario-based Bayesian Network Reasoning.”
Reviewing the impact of social media on emergency management
Tapia's presentation, "Developing for the Last Mile of Crisis: Where Analytics End and Community and Organization Begin," was one of three keynote addresses delivered throughout the week. Her research seeks to develop information and communication technology solutions that promote better decision-making across all emergency responders by using data generated via mobile information technologies.
Along with Tapia, much of the IST group’s research centered on the use of social media and online information to facilitate emergency management and response. For example, barriers to using citizen-reported information — information overload, lack of trust in available information, and lack of staff to moderate the tools — provided the guiding questions for Grace, Kropczynski and Tapia in their paper, “Community Coordination: Aligning Social Media Use in Community Emergency Management.”
“We suggest that these barriers can be overcome by aligning social media use within existing incident management systems,” the group wrote in their paper. “In particular, our findings recommend integrating social media capabilities within Public Service Answering Points, 9-1-1 call centers that already process citizen-reported information, and coordinating inter-agency and public incident notification and emergency dispatch.”
Noting that these resources are professionally staffed and follow trusted protocols, the group proposed that social media integration into existing response operations can allow emergency managers to better “detect, verify and dispatch information reported on social media using existing sources of official and citizen-reported information.”
Sharing best practices for scenario-based emergency training
Graham and Mark B. Stephens, professor of family and community medicine at Penn State College of Medicine, shared best practices and future directions for scenario-based emergency training through their paper, “Analytic Decision Gaming — A Tool to Develop Crisis Response and Clinical Reasoning.”
Serving as the coordinator for IST’s security and risk analysis undergraduate program, Graham has implemented analytic decision games into the SRA curriculum. These exercises provide unique crises that require each participant to adopt a different role, such as acting as the adversary or taking the lead as an emergency responder, to effectively manage the crisis.
“Red cell teaming develops new strategies; exposes biases emerging from groupthink; identifies gaps and vulnerabilities that are overlooked because of over familiarity with the operation and/or security environment; highlights areas for improved inter-organizational cooperation in preparedness, response and mitigation; and provides recommendations for insuring unchecked assumptions do not become threats,” the pair wrote in their paper.
In the past two years, Graham and Stephens have adapted this semester-long classroom exercise into a four-day training for students in the Penn State College of Medicine. The scenario tasks participants with identifying and resolving issues after the complete failure of the U.S. power grid and a series of localized emergencies and compounding countywide failures.
In their paper, Graham and Stephens offer ways to refine the exercise for future use and adapt it for different participant groups.
“The time-constrained nature of the medical school version fosters a sense of priority and focus, which better emulates conditions under stress as may be found in a real crisis event,” they explained. “[The analytic decision game] is the perfect backdrop for medical students to develop the clinical reasoning skills required to address the needs of the local community and build a foundation for transferring these skills to a national and global context.”
Exploring how unverified breaking news influences crisis response
Lalone, Kropczynski and Tapia also explored the ways existing social media and online information platforms are adjusting to the spread of breaking, and often unverified, news. Their paper, "The Symbiotic Relationship of Crisis Response Professionals and Enthusiasts as Demonstrated by Reddit’s User-Interface Over Time,” aimed to understand how the actions and media coverage of crisis response professionals influence actual response behavior.
To do this, the group looked at Reddit's response and redesign after the site faced criticism after the Boston Marathon Bombing in 2013. After site users incorrectly attempted to identify the bombers, there was a clash between crisis response professionals and the public about the role of crowdsourcing public response.
The group also studied the veracity of publicly submitted information for other crises, including the Aurora, Colorado, movie theater shooting and the unrest in Ferguson, Missouri, to understand the challenges surrounding information verification and the spread of online rumors. With Reddit users shifting the platform from news aggregator to a breaking news outlet, the platform sought to address issues of power and abuse by redesigning and redeveloping its user interface.
“While these efforts have yet to officially become representative of crisis response professionals, spaces and communities like Reddit are providing more reliable information with each response," the group stated in their research. “Reddit is constantly re-working its software to define what matters to information flow within a crisis. It also defines what is important for crisis response professionals to do and what information verification requires.”
Throughout their presentations, IST researchers continued to highlight how emerging technologies play a role in crisis response. Shifting these technologies from community tools to helpful resources is a deliberate, but necessary, change.
“The key to regular use of social media data in emergency response is making it normal, common, [and used] every day for the responders and their support organizations,” Tapia said in her keynote speech. “People reach for the tools they know in a crisis. We believe that by inserting knowledge derived from the community into existing systems at the most mundane, everyday emergencies, we can slowly nudge a whole system.”