127
click to vote
ACL
2012
13 years 5 months ago
2012
This paper presents an unsupervised approach to learning translation span alignments from parallel data that improves syntactic rule extraction by deleting spurious word alignment...