Survey of Event Detection in Twitter

PhD Qualifying Examination

Title: "Survey of Event Detection in Twitter"




This survey reviews event detection researches using Twitter data, which 
have become a popular research topic for social media platforms in the 
area of information retrieval. Though event detection methods have long 
been studied, but most of the traditional event detection works focus on 
news document sets. The popularity of social media platforms like Twitter 
brings new opportunity and challenges to this traditional task. Due to the 
high volume, short length, unstructured and streaming nature of micro blog 
posts. We first provide a general review of the popular research topics in 
Twitter. Secondly, we summarize and categorize several widely adopted 
techniques in event detection in Twitter and introduce some highly cited 
popular works. It is non-trivial to conduct the evaluation of event 
detection in Twitter, as the ground truth is not straight forward. Then we 
introduced some widely accepted evaluation methods for event detection in 
Twitter. Lastly we focus on the latest trend of event detection, which is 
leveraging Twitter as a crow sensor to detect natural disasters, and 
introduce some representative works.

Date:			Monday, 11 December 2017

Time:                  	9:30am - 11:30am

Venue:                  Room 5508
                         Lifts 25/26

Committee Members:	Dr. Wilfred Ng (Supervisor)
 			Prof. Lei Chen (Chairperson)
 			Dr. Qiong Luo
 			Dr. Raymond Wong

**** ALL are Welcome ****