Parthasarathy team receives NSF funds to study social sensors in disaster responses

Changing the Disaster Narrative

The story of disaster management is typically one of failure. When catastrophes such as floods or hurricanes strike, the aftereffects include structural failures, power failures, transportation failures and communication failures.

A team led by TDA@OhioState affiliate Srinivasan Parthasarathy, PhD, is working to change that narrative. Funded by an award through the National Science Foundation’s Interdisciplinary Research in Hazards and Disasters program, the group will work to find new ways of assessing and addressing damage caused by disasters. By combining data from physical and social sensors, they will help plot a path that helps disaster recovery teams act more effectively and efficiently.

The team is a transdisciplinary group. Principal Investigator Parthasarathy comes from the Department of Computer Science and Engineering, with expertise in data analytics and biomedical informatics as director of the Data Mining Research Laboratory. He is joined by Desheng Liu, PhD, an expert in flood mapping from the Department of Geography, and Ethan Kubatko, PhD, from the Department of Civil, Environmental and Geodetic Engineering, who has expertise in storm surge modeling.

The Ohio State trio will work on the project with a team from Wright State University led by Ohio State alumnus Amit Sheth, Director of the Ohio Center of Excellence in Knowledge-enabled Computing, studying semantic web technologies, and his colleagues Valerie Shalin, PhD, representing the social sciences and psychology, and T.K. Prasad, PhD, who brings with him expertise in sensors and traffic modeling.

Together, the researchers will investigate ways for recovery teams to gather and analyze data in the wake of a natural disaster, requiring fundamental advances in real-time sensor integration, semantic and network analysis, modeling, and flood mapping.

An exemplar challenge the team seeks to solve, posed by Parthasarathy, is the following: “What are the cascade effects of a disaster across multiple networked infrastructure components that form the fabric of modern society?” Experience with disasters such as Hurricanes Katrina and Sandy suggest some preliminary exemplar answers to the question.

“A sudden storm surge could lead to the breaking of crucial levy (civil infrastructure), which in turn could lead to flooding (reported by citizen sensors and physical sensors), which in turn can lead to a substation outage (power grid), which in turn can impact traffic flow (road network),” Parthasarathy says. “Understanding, integrating and automatically extracting such cascade patterns across layers of infrastructure can lead to game-changing advances in disaster response.”

While physical sensors have been employed for decades to help mitigate and guide disaster response, social sensors are newer tools that can help fill information gaps for recovery teams. Monitoring citizen reports on social media can give responders real time information from places where physical sensors are absent. Parthasarathy and his group will create a means to integrate data from both social and physical sensors, and use it to guide disaster recovery efforts in the wake of cascade effects.

Of the 136 disaster management research proposals received by the National Science Foundation, Ohio State’s is one of only 11 to receive funding. To date, the team has been awarded $1,975,000. When its student research component is completed, the award is expected to total more than $2 million.

Suggested Articles
User-generated data is a social science goldmine

As increasing amounts of digital data are produced and stored online it is important to remember that humans produce much of that data. In an era in which people express...

HackOHI/O 2016 connects companies and students

HackOHI/O 2016 has mentoring and sponsorship opportunities for organizations interested in participating in Ohio State’s student-run hackathon, which will be held Nov. 19-20, 2016. Sponsor benefits include serving as a judge, hosting a challenge, demonstrating...

Nandi receives IEEE Early Career Award, credits collaborators

Arnab Nandi The notion of a database search without a keyword sounds paradoxical. How can you search for something you can’t name? TDA affiliate Arnab Nandi, PhD, Assistant Professor in Computer...

Visual analytics for management: Translational science and applications in practice

An annotated excerpt from the forthcoming text by Elliot Bendoly and Sasha Clark; Taylor & Francis, 2017  Amidst a data ecosystem’s collectors, analysts and users, important meaning can be distorted...

Existing data yield insights into depression among understudied groups

Macarius Donneyong’s research involves digging into existing datasets to find golden nuggets of insight that can improve healthcare outcomes in real-world situations. An assistant professor of pharmacy practice and science...