Event Detection over Twitter Social Media Streams
Speaker: Dr Xiangmin (Emily) Zhou
Date: Friday 20 December 2013
Time: 4:00pm
Location:Blackfriars Campus CC05.GD.03
Seminar Chairman: Professor Longbing Cao, Director, Advanced Analytics Institute (AAi)
Abstract:
Microblogs have recently become an important source for reporting real-world events. Social events may hold critical materials that describe the situations during a crisis. In real applications, such as crisis management and decision making, monitoring the critical events over social streams will enable watch officers to analyze a whole situation and make the right decision based on the contexts like what is happening, where it is, who are involved, and its effects etc. However, due to the special characteristics of social data, such as highly uncertain, highly complex and huge amount, effectively and efficiently monitoring events over social streams has been challenging. In this talk, I will present a social event detection framework, which exploits the information of social data over multiple dimensions, and considers the special characteristics of social data. Specially, we first propose a graphical topic model to capture the content, time, and location of messages. Then the similarity between two messages is measured by the distance between their topic distributions. We further enhance the similarity by embedding the social linkage difference. Finally, the events are identified by conducting efficient similarity joins over social media streams. We conduct extensive experiments over data from Twitter during two disasters, Cyclone ULUI and flooding in Queensland from 2010 to 2011, to prove the high effectiveness and efficiency of the proposed approach.
Biography:
Xiangmin Zhou received her PhD in Computer Science from the University of Queensland (Australia) in 2008. She worked in CSIRO as a Postdoctoral Research Fellow from June 2008 to June 2011, and as a Research Scientist from July 2011 to Dec. 2012. She has been an Adjunct Research Fellow at Australian National University since May 2012. Her current research interests lie in social media/network analysis and mining, multimedia database and streams, cloud computing. Her work has been published in top conferences and top journals, including ACM Multimedia Conference, ICDE Conference, VLDB Journal, IEEE Trans. on Multimedia, IEEE Trans. on Knowledge and Data Engineering etc. She won the best student paper award at Australian Database Conference 2003. She has been involved as a PC member in a number of conferences, including ICDE, ACM Multimedia, ASONAM, ICDKE etc. She has served as an invited reviewer of many journals, such as VLDB Journal, IEEE Trans. on Multimedia, Multimedia System, WWW Journal and IEEE TKDE etc.