BEGIN:VCALENDAR
VERSION:2.0
PRODID:-//Talks.cam//talks.cam.ac.uk//
X-WR-CALNAME:Talks.cam
BEGIN:VEVENT
SUMMARY:Bursty and Hierarchical Structure in Streams. - Philip Sterne (Uni
 versity of Cambridge)
DTSTART:20090828T100000Z
DTEND:20090828T110000Z
UID:TALK19527@talks.cam.ac.uk
CONTACT:Philip Sterne
DESCRIPTION:The high-level idea is to analyze a stream of documents and fi
 nd features whose behavior is `bursty': they occur with high intensity ove
 r a limited period of time. The analysis uses a probabilistic automaton wh
 ose states correspond to the frequencies of individual words. For each wor
 d separately\, one imagines a copy of the automaton generating occurrences
  of the word\, and computes its most likely state sequence over the course
  of the stream. State transitions correspond to points in time around whic
 h the frequency of the word changes significantly -- that is\, to the begi
 nning or end of a `burst' in the usage of the word. The full details of al
 l this are given in the paper "here":http://www.cs.cornell.edu/home/kleinb
 er/bhs.pdf (also have a look at some of the "results":http://www.cs.cornel
 l.edu/home/kleinber/kdd02.html )
LOCATION:TCM Seminar Room\, Cavendish Laboratory\, Department of Physics
END:VEVENT
END:VCALENDAR