Spatiotemporal mining in microblogs

PhD Qualifying Examination


Title: "Spatiotemporal mining in microblogs"

by

Mr. Eleftherios NTAFLOS


Abstract:

Recently, microblogging services are getting increasing attention from the 
research community. A main cause of this trend is their inherent spatial and 
temporal information, which can find several uses, such as identifying local 
trends and events, detecting emergency situations, predicting and monitoring 
natural phenomena, etc. The growing demand of location based services for the 
purpose of user profiling, marketing, advertising and recommending, adds even 
more applications to this movement. In this survey we group microblog 
spatiotemporal mining into three main categories: i) user location estimation, 
ii) topic, event, and trend detection, and iii) correlation between topics and 
locations. The state of the art approaches for those problems span the fields 
of natural language processing, information retrieval, machine learning, text 
mining and database management. We present and analyze several of those 
techniques, in order to gain a better understanding of their underlying 
mechanisms.


Date:			Friday, 13 December 2013

Time:                   2:00pm - 4:00pm

Venue:                  Room 3311
                         lifts 17/18

Committee Members:	Prof. Dimitris Papadias (Supervisor)
 			Prof. Frederick Lochovsky (Chairperson)
 			Dr. Qiong Luo
 			Dr. Raymond Wong


**** ALL are Welcome ****