Detecting Help Seeking Tweets in Natural Crises
October 19, 2016
Track: ACM SRC
Help seeking posts on social media contain valuable actionable information but are difficult to be detected among millions of posts. Most of the detection methods exploit textual features and usefulness of the context (such as location) has been overlooked. In this poster, we use context to propose a method for detecting help seeking tweets and provide experimental results on Hurricane Sandy data.