BEGIN:VCALENDAR
VERSION:2.0
PRODID:-//Talks.cam//talks.cam.ac.uk//
X-WR-CALNAME:Talks.cam
BEGIN:VEVENT
SUMMARY:Interactive Data Mining - Towards Mixed Initiative Approaches - Ma
 rtin Spott\, BT Research and Innovation
DTSTART:20130214T111500Z
DTEND:20130214T121500Z
UID:TALK43297@talks.cam.ac.uk
CONTACT:Alistair Stead
DESCRIPTION:When something goes wrong in an organisation\, analysts and do
 main experts work hand in hand to find root causes\, led by their experien
 ce and beliefs. In other words\, they typically only find what they are lo
 oking for. Successful root cause analyses therefore require creativity\, t
 ime and a portion of luck.  \n\nModern data mining techniques address this
  problem by exploring large search spaces in a short time\, producing a se
 t of potentially interesting patterns. However\, machines cannot judge how
  interesting a pattern is. Analysts may find the results disappointing\, b
 ecause they still have to go through a large number of patterns\, some of 
 which they will know and others they are not interested in. \n\nThe presen
 tation will introduce our own techniques for pattern mining over time and 
 address the following questions: \n\n# How can a machine judge how interes
 ting a pattern is?\n# How do we include time as a dimension?\n# How can we
  present\, visualise and explore data mining results?\n# What would a trul
 y interactive data mining system look like that combines the speed of mach
 ine computation with the experience and domain expertise of analysts?  
LOCATION:Rainbow Room (SS03)\, Computer Laboratory
END:VEVENT
END:VCALENDAR