Paper: Empirical lower bounds on translation unit error rate for the full class of inversion transduction grammars

ACL ID W09-3805
Title Empirical lower bounds on translation unit error rate for the full class of inversion transduction grammars
Venue International Conference on Parsing Technologies
Session Main Conference
Year 2009
Authors
  • Anders Søgaard (University of Copenhagen, Copenhagen Denmark)
  • Dekai Wu (University of Science and Technology, Clear Water Bay Hong Kong)

Empirical lower bounds studies in which the frequency of alignment configurations that cannot be induced by a particular for- malism is estimated, have been important for the development of syntax-based ma- chine translation formalisms. The for- malism that has received most attention has been inversion transduction grammars (ITGs) (Wu, 1997). All previous work on the coverage of ITGs, however, con- cerns parse failure rates (PFRs) or sen- tence level coverage, which is not di- rectly related to any of the evaluation mea- sures used in machine translation. Søgaard and Kuhn (2009) induce lower bounds on translation unit error rates (TUERs) for a number of formalisms, incl. normal form ITGs, but not for the full class of ITGs. Many of the alignment configurations that cannot be induced by n...