BEGIN:VCALENDAR
VERSION:2.0
PRODID:-//Talks.cam//talks.cam.ac.uk//
X-WR-CALNAME:Talks.cam
BEGIN:VEVENT
SUMMARY:What is happening now? Finding events in Massive Message Streams -
  Miles Osborne\, University of Edinburgh School of Informatics
DTSTART:20101110T141500Z
DTEND:20101110T151500Z
UID:TALK26387@talks.cam.ac.uk
CONTACT:Stephen Clark
DESCRIPTION:Social Media {eg Twitter\, Blogs\, Forums\, FaceBook} has expl
 oded over\nthe last few years. FaceBook is now the  most visited site on t
 he Web\,\nwith Blogger being the 7th and Twitter the 13th. These sites con
 tain\nthe aggregated beliefs and opinions of millions of people on an epic
 \nrange of topics\, and in a large number of languages. Twitter in\npartic
 ular is an example of a massive message stream and finding\nevents embedde
 d in it poses hard engineering challenges.  I will\nexplain how we use a v
 ariant of Locality Sensitive Hashing to find new\nstories as they break. T
 he approach scales well\, easily dealing with\nthe more than 1 million Twe
 ets a day we process and only needing a\nsingle processor. For June 2009\,
  the fastest growing stories all\nconcerned deaths of one kind or another.
 \n\nBio: Miles Osborne is a Reader in Informatics\, Edinburgh.  His\nlong-
 standing reseach interests include machine translation and\nalgorithms for
  dealing with massive datasets.  More recently his\ninterests have extende
 d to Social Media.\n
LOCATION:Lecture Theatre 1\, Computer Laboratory
END:VEVENT
END:VCALENDAR