Natural Language Processing to the Rescue? Extracting "Situational Awareness" Tweets during Mass Emergency thumbnail
Pause
Mute
Subtitles
Playback speed
0.25
0.5
0.75
1
1.25
1.5
1.75
2
Full screen

Natural Language Processing to the Rescue? Extracting "Situational Awareness" Tweets during Mass Emergency

Published on Aug 18, 20114141 Views

In times of mass emergency, vast amounts of data are generated via computer-mediated communication (CMC) that are difficult to manually cull and organize into a coherent picture. Yet valuable infor

Related categories

Chapter list

Natural Language Processing to the Rescue?00:00
Microblogging & Emergency00:16
Twitter & Emergency00:35
Situational Awareness01:50
Tweets during Mass Emergencies02:25
Tweets with Information Contributing to Situational Awareness (1)03:26
Tweets with Information Contributing to Situational Awareness (2)03:45
Tweets with Information Contributing to Situational Awareness (3)03:51
Hypothesis03:57
Goals04:19
Our Approach (1)04:38
Our Approach (2)04:46
Our Approach (3)04:53
Our Approach (4)05:00
Datasets05:10
2009 & 2010 Red River Floods05:38
2009 Oklahoma Fires06:12
2010 Haiti Earthquake06:41
Annotation06:57
Situational Awareness (SA) (1)07:32
Situational Awareness (SA) (2)08:10
Subjectivity (1)08:27
Subjectivity (2)09:02
Register (1)09:27
Register (2)10:01
Style (1)10:21
Style (2)11:01
Classification11:18
Classification -Technique11:45
Classification –Training data12:10
Classification - Overview12:19
Simple Features13:01
Linguistic Features13:34
Classification Results (1)14:06
SA Classification using Naïve Bayes and MaxEnt with all features14:21
Classification Results (2)15:10
MaxEnt Results on Balanced Dataset (1)15:39
MaxEnt Results on Balanced Dataset (2)16:39
MaxEnt Results on Balanced Dataset (3)16:56
MaxEnt Results on Balanced Dataset (4)17:19
MaxEnt Results on Balanced Dataset (5)17:27
MaxEnt Results on Balanced Dataset (6)17:50
MaxEnt Results on Balanced Dataset (7)17:57
MaxEnt Results on Balanced Dataset (8)18:04
Classification Results (3)18:09
MaxEnt Results on Balanced Dataset (9)18:24
MaxEntResults across all Events18:29
Conclusions19:52
Acknowledgements20:19