Information for disaster relief from Twitter feed analysis
Brown Bag Lecture by Dr. Jill Crisman | 12/12/2017 11:00AM – 12PM | 7th Floor Conference Room, Bldg 38A
The Text-enabled Humanitarian Operations Runtime (THOR) system automatically processes text and speech messages to support Humanitarian Aid and Disaster Relief applications as part of the DARPA Low Resource Languages for Emergent Incidents (LORELEI) program. This program is experimenting with processes to quickly derive models for low-resource languages (e.g., Uyghur), and implementing the models in algorithms that automatically interpret text or speech from these languages. THOR employs these algorithms to help users ascertain the situation on the ground during a relief effort. We will describe the types of algorithms being developed, show some typical automated processing results, and demonstrate how we use these results in our application.
As the Chief Scientist at Next Century Corporation, Dr. Crisman is responsible for technology leadership and execution of strategic technical initiatives. Her research and expertise include computer vision, robotics, and machine learning. She was one of the founding faculty members at the Franklin W. Olin College of Engineering, and also served as an Associate Professor and Director of the Robotics and Vision System Laboratory at Northeastern University. Dr. Crisman earned her Ph.D. from Carnegie Mellon University in Electrical and Computer Engineering.