Reliable Extraction of Emergency Response Networks from Text Data and Benchmarking with National Emergency Response Guidelines

This project employs techniques from natural language processing and social network analysis to identify and evaluate multi-modal networks involved in Humanitarian Assistance and Disaster Relief (HADR) efforts. Empirical evidence derived from text data will be compared to the expected response behavior as set out in national guidelines, such as Department of Homeland Security's National Response Framework, to identify congruence and opportunities and needs for policy change.

The project makes both empirical and methodological contributions to current standards and practices: (1) empirical: we evaluate the effectiveness of current response frameworks in terms of facilitating inter-organizational collaboration structures among local, state, federal, non-governmental entities. (2) methodological: we determine best practices for constructing reliable semantic networks from large-scale text data, as well as building a domain-specific (HADR-specific) classification schema of network edges that can be used in future HADR-related networks.