BEGIN:VCALENDAR
VERSION:2.0
PRODID:-//Talks.cam//talks.cam.ac.uk//
X-WR-CALNAME:Talks.cam
BEGIN:VEVENT
SUMMARY: Three Stories on Aggregated Search - Maarten de Rijke\, Universit
 y of Amsterdam
DTSTART:20131120T160000Z
DTEND:20131120T170000Z
UID:TALK47971@talks.cam.ac.uk
CONTACT:Tamara Polajnar
DESCRIPTION:A result page of a modern search engine is often much more com
 plicated than a simple list of "ten blue links.” In particular\, a searc
 h engine may combine results from different sources (e.g.\, web\, news\, i
 mages\, structured data)\, and display these as grouped results to provide
  a better user experience. Such a system is called an aggregated or federa
 ted search system.\n\nIn the talk I will touch on three aspects of aggrega
 ted search. First\, I will talk about aggregated search for queries with a
  video intent\, where multiple sources being aggregated might have results
  of the same (video) type. Next\, I will illustrate new challenges for ent
 ity linking and disambiguation that emerge in the setting of aggregated se
 arch. And finally\, I will talk about evaluating aggregated search. Becaus
 e search engines evolve over time\, their results need to be constantly ev
 aluated. However\, one of the most efficient and widely used evaluation me
 thods\, interleaving\, cannot be directly applied to aggregated search sys
 tems\, as it ignores the need to group results originating from the same s
 ource (vertical results).\n\nThe talk is based on joint work with Marc Bro
 n\, Aleksandr Chuklin\, Katja Hofmann\, Damien Lefortier\, Daan Odijk\, An
 ne Schuth\, Pavel Serdyukov\, Jasmijn Van Gorp
LOCATION:SW 01\, Computer Laboratory\, William Gates Builiding
END:VEVENT
END:VCALENDAR