BEGIN:VCALENDAR
VERSION:2.0
PRODID:-//Talks.cam//talks.cam.ac.uk//
X-WR-CALNAME:Talks.cam
BEGIN:VEVENT
SUMMARY:An Operation Sequence Model for Explainable Neural Machine Transla
 tion - Felix Stahlberg\, CUED\, University of Cambridge
DTSTART:20190614T110000Z
DTEND:20190614T120000Z
UID:TALK121996@talks.cam.ac.uk
CONTACT:Andrew Caines
DESCRIPTION:We propose to achieve explainable neural machine translation (
 NMT) by changing the output representation to explain itself. We present a
  novel approach to NMT which generates the target sentence by monotonicall
 y walking through the source sentence. Word reordering is modeled by opera
 tions which allow setting markers in the target sentence and move a target
 -side write head between those markers. In contrast to many modern neural 
 models\, our system emits explicit word alignment information which is oft
 en crucial to practical machine translation as it improves explainability.
  Our technique can outperform a plain text system in terms of BLEU score u
 nder the recent Transformer architecture on Japanese-English and Portugues
 e-English\, and is within 0.5 BLEU difference on Spanish-English. 
LOCATION:FW26\, Computer Laboratory
END:VEVENT
END:VCALENDAR