BEGIN:VCALENDAR
VERSION:2.0
PRODID:-//Talks.cam//talks.cam.ac.uk//
X-WR-CALNAME:Talks.cam
BEGIN:VEVENT
SUMMARY:Unsupervised Question Answering - Patrick Lewis\, UCL
DTSTART:20191024T100000Z
DTEND:20191024T110000Z
UID:TALK132967@talks.cam.ac.uk
CONTACT:Edoardo Maria Ponti
DESCRIPTION:Obtaining training data for Question Answering (QA) is time-co
 nsuming and costly\, and existing QA datasets are only available for limit
 ed domains and languages. In this talk\, we'll explore to what extent high
  quality training data is actually required for Extractive QA\, and invest
 igate the possibility of unsupervised Extractive QA. We approach this prob
 lem by first learning to generate context\, question and answer triples in
  an unsupervised manner\, which we then use to synthesize Extractive QA tr
 aining data automatically.  We find that modern QA models can learn to ans
 wer human questions surprisingly well using only synthetic training data. 
 We demonstrate that\, without using the SQuAD training data at all\, our a
 pproach achieves 56.4 F1 on SQuAD v1 (64.5 F1 when the answer is a Named e
 ntity mention)\, outperforming early supervised models.\nWe will also expl
 ore methods to build cross-lingual Question Answering models which do not 
 require cross-lingual supervision (zero-shot language transfer)\, as well 
 as the challenge of how to fairly evaluate their performance in many targe
 t languages. 
LOCATION:Board room\, Faculty of English\, 9 West Rd (Sidgwick Site)
END:VEVENT
END:VCALENDAR