Intelligent Analysis of Big Social Media Data for Crisis Tracking


Often, a well-organized, concerted response can greatly reduce the damage inflicted by large-scale disasters. An important factor affecting our ability to efficiently and promptly respond to crises is the availability of quality information to first-responders about the ground realities of the affected areas. During times of crises, conventional sources of communication are often disrupted, and hence, the availability and reliability of information are called into question. However, we can leverage the vast amounts of data from social media sources such as Twitter to provide relevant information on resources and aid required in the affected areas. The project is aimed at shifting the focus of the applications of big social media data from simple tracking and statistics to a principled set of methods for intelligent analysis of data to provide situational and effective insights for crisis tracking, response planning, humanitarian aid, and disaster relief. As the collective data collection capability has improved significantly, the pressing need for intelligent analysis arises to overcome problems associated with big social media data: information overload, misinformation, covert adversaries, false positives, and differential information discovery. Extending the successes of these software systems for the tracking of blog, tweets and information gathering, this proposed project aims to expand data analysis that can intelligently handle information overload and misinformation, and intelligently differentiate novel information among big social media data from information of known sources. This project has three major research objectives for the intelligent analysis of big social media data:
This project has three major research objectives for the intelligent analysis of big social media data:


Our latest system TweetTracker focuses on tweets collected from Twitter to support event monitoring and analysis. More information about this project can be found here.

Conference Tutorials

Invited Talks


  • Awards

  • Project Members (current and former)


    This project is sponsored by ONR grant N00014-16-1-2257

    Created on March 13, 2017
    Contact: Huan Liu.

    Last Updated: April 11, 2018