BEGIN:VCALENDAR
VERSION:2.0
PRODID:-//Talks.cam//talks.cam.ac.uk//
X-WR-CALNAME:Talks.cam
BEGIN:VEVENT
SUMMARY:NLIP reading group: Connecting the Dots Between News Articles - Di
 armuid Ó Séaghdha (University of Cambridge)
DTSTART:20110512T110000Z
DTEND:20110512T120000Z
UID:TALK31126@talks.cam.ac.uk
CONTACT:Jimme Jardine
DESCRIPTION:Diarmuid will be kicking off this term's reading group series 
 with the following paper:\n\n@conference{shahaf2010connecting\,\n  title={
 {Connecting the dots between news articles}}\,\n  author={Shahaf\, D. and 
 Guestrin\, C.}\,\n  booktitle={Proceedings of the 16th ACM SIGKDD internat
 ional conference on Knowledge discovery and data mining}\,\n  pages={623--
 632}\,\n  year={2010}\,\n  organization={ACM}\n}\n\nThe process of extract
 ing useful knowledge from large datasets has become one of the most pressi
 ng problems in today's society. The problem spans entire sectors\, from sc
 ientists to intelligence analysts and web users\, all of whom are constant
 ly struggling to keep up with the larger and larger amounts of content pub
 lished every day. With this much data\, it is often easy to miss the big p
 icture.\n\nIn this paper\, we investigate methods for automatically\nconne
 cting the dots { providing a structured\, easy way to navigate within a ne
 w topic and discover hidden connections. We focus on the news domain: give
 n two news articles\, our system automatically nds a coherent chain linki
 ng them together. For example\, it can recover the chain of events startin
 g with the decline of home prices (January 2007)\, and ending with the ong
 oing health-care debate.\n\nWe formalize the characteristics of a good cha
 in and provide an ecient algorithm (with theoretical guarantees) to conne
 ct two xed endpoints. We incorporate user feedback\ninto our framework\, 
 allowing the stories to be rened and personalized. Finally\, we evaluate 
 our algorithm over real news data. Our user studies demonstrate the algori
 thm's\neectiveness in helping users understanding the news.
LOCATION:GS15\, Computer Laboratory
END:VEVENT
END:VCALENDAR
