Actions

DisasterAnalytics: Difference between revisions

Computational Linguistics and Information Processing

No edit summary
No edit summary
Line 1: Line 1:
* The ubiquitous presence of cell phones and social media present new opportunities to analyze not only human mobility but also online citizens’ communications. This project focuses on the use of large-scale, georeferenced data extracted from ubiquitous technologies to model both human mobility, to understand how humans respond to shocks, and online communications, to provide hypothesis that could explain the observed responses. For that purpose, we combine data mining and machine learning techniques to reliably characterize and predict mobility and communication patterns during shocks with the objective of helping decision makers and emergency responders to improve preparedness and response policies.
The ubiquitous presence of cell phones and social media present new opportunities to analyze not only human mobility but also online citizens’ communications. This project focuses on the use of large-scale, georeferenced data extracted from ubiquitous technologies to model both human mobility, to understand how humans respond to shocks, and online communications, to provide hypothesis that could explain the observed responses. For that purpose, we combine data mining and machine learning techniques to reliably characterize and predict mobility and communication patterns during shocks with the objective of helping decision makers and emergency responders to improve preparedness and response policies.
 
* In this project, we develop two research thrusts. The first one focuses on the development of a novel framework to analyze behavioral changes during disasters using Call Detail Records (CDRs) from a telecommunications company. CDR datasets are collections of spatio-temporal traces that can characterize individual mobility and social network behaviors at very fine scales. The proposed framework exploits the granular behavioral models to evaluate the similarities and differences between normal and response patterns observed during shocks. Such framework, was used to analyze Rwanda’s 2012 floods and show that disasters tend to disrupt both mobility patterns and communication behaviors while recovery times can take several weeks.  
 
* The second thrust explores  a semi-automatic framework to extract and compare, in retrospect, the digital communication footprints of citizens and governments during disasters. These footprints, which characterize the topics discussed during a disaster at different spatio-temporal scales, are computed in an unsupervised manner using topic models, and manually labeled to identify specific issues affecting the population. The end objective is to offer detailed information about issues affecting citizens during natural disasters and to compare these against local governments’ communications. Given that a growing number of citizens and local governments have embraced the use of Twitter to communicate during natural disasters, we evaluate the framework using Twitter communications from 18 snowstorms (including two blizzards) on the US east coast.
In this project, we develop two research thrusts. The first one focuses on the development of a novel framework to analyze behavioral changes during disasters using Call Detail Records (CDRs) from a telecommunications company. CDR datasets are collections of spatio-temporal traces that can characterize individual mobility and social network behaviors at very fine scales. The proposed framework exploits the granular behavioral models to evaluate the similarities and differences between normal and response patterns observed during shocks. Such framework, was used to analyze Rwanda’s 2012 floods and show that disasters tend to disrupt both mobility patterns and communication behaviors while recovery times can take several weeks.  
 
The second thrust explores  a semi-automatic framework to extract and compare, in retrospect, the digital communication footprints of citizens and governments during disasters. These footprints, which characterize the topics discussed during a disaster at different spatio-temporal scales, are computed in an unsupervised manner using topic models, and manually labeled to identify specific issues affecting the population. The end objective is to offer detailed information about issues affecting citizens during natural disasters and to compare these against local governments’ communications. Given that a growing number of citizens and local governments have embraced the use of Twitter to communicate during natural disasters, we evaluate the framework using Twitter communications from 18 snowstorms (including two blizzards) on the US east coast.


Video: https://www.youtube.com/watch?v=o6i0ff_sYYo&feature=youtu.be
Video: https://www.youtube.com/watch?v=o6i0ff_sYYo&feature=youtu.be

Revision as of 02:32, 9 December 2017

The ubiquitous presence of cell phones and social media present new opportunities to analyze not only human mobility but also online citizens’ communications. This project focuses on the use of large-scale, georeferenced data extracted from ubiquitous technologies to model both human mobility, to understand how humans respond to shocks, and online communications, to provide hypothesis that could explain the observed responses. For that purpose, we combine data mining and machine learning techniques to reliably characterize and predict mobility and communication patterns during shocks with the objective of helping decision makers and emergency responders to improve preparedness and response policies.   In this project, we develop two research thrusts. The first one focuses on the development of a novel framework to analyze behavioral changes during disasters using Call Detail Records (CDRs) from a telecommunications company. CDR datasets are collections of spatio-temporal traces that can characterize individual mobility and social network behaviors at very fine scales. The proposed framework exploits the granular behavioral models to evaluate the similarities and differences between normal and response patterns observed during shocks. Such framework, was used to analyze Rwanda’s 2012 floods and show that disasters tend to disrupt both mobility patterns and communication behaviors while recovery times can take several weeks.

The second thrust explores a semi-automatic framework to extract and compare, in retrospect, the digital communication footprints of citizens and governments during disasters. These footprints, which characterize the topics discussed during a disaster at different spatio-temporal scales, are computed in an unsupervised manner using topic models, and manually labeled to identify specific issues affecting the population. The end objective is to offer detailed information about issues affecting citizens during natural disasters and to compare these against local governments’ communications. Given that a growing number of citizens and local governments have embraced the use of Twitter to communicate during natural disasters, we evaluate the framework using Twitter communications from 18 snowstorms (including two blizzards) on the US east coast.

Video: https://www.youtube.com/watch?v=o6i0ff_sYYo&feature=youtu.be