BEGIN:VCALENDAR
VERSION:2.0
PRODID:-//Talks.cam//talks.cam.ac.uk//
X-WR-CALNAME:Talks.cam
BEGIN:VEVENT
SUMMARY:Social Media Predictive Analytics: Methods and Applications - Svit
 lana Volkova\, Johns Hopkins University
DTSTART:20150317T100000Z
DTEND:20150317T110000Z
UID:TALK58383@talks.cam.ac.uk
CONTACT:Microsoft Research Cambridge Talks Admins
DESCRIPTION:Large-scale real-time social media analytics provides a novel 
 set of conditions for the construction of predictive models. With individu
 al users as training and test instances\, their associated content (“lex
 ical features”) and context ("network features") are made available incr
 ementally over time\, as they converse over discussion forums. We propose 
 various approaches to handling this dynamic data for predicting latent use
 r properties\, from traditional batch training and testing\, to incrementa
 l bootstrapping\, and then active learning via interactive rationale crowd
 sourcing.\n\nWe also study the relationships between a variety of predicte
 d user properties\, opinions and emotions on a large sample of users in on
 line social network. We first correlate user demographics and personality 
 with the emotional profile emanating from user tweets. We then analyze the
  relationships between predicted user properties and user-environment emot
 ional contrast estimated over various neighborhoods including friends\, re
 tweeted and mentioned users. Finally\, we analyze and compare predictive p
 ower of latent user properties\, emotions and interests for automatically 
 inferring showing off and self-promoting behaviors projected in online soc
 ial networks.
LOCATION:Small Lecture Theatre\, Microsoft Research Ltd\, 21 Station Road\
 , Cambridge\, CB1 2FB
END:VEVENT
END:VCALENDAR