Paper: Analysis and Prediction of Unalignable Words in Parallel Text

ACL ID E14-4037
Title Analysis and Prediction of Unalignable Words in Parallel Text
Venue Annual Meeting of The European Chapter of The Association of Computational Linguistics
Session Main Conference
Year 2014
Authors

Professional human translators usually do not employ the concept of word align- ments, producing translations ?sense-for- sense? instead of ?word-for-word?. This suggests that unalignable words may be prevalent in the parallel text used for ma- chine translation (MT). We analyze this phenomenon in-depth for Chinese-English translation. We further propose a sim- ple and effective method to improve au- tomatic word alignment by pre-removing unalignable words, and show improve- ments on hierarchical MT systems in both translation directions. 1 Motivation It is generally acknowledged that absolute equiva- lence between two languages is impossible, since concept lexicalization varies across languages. Major translation theories thus argue that texts should be translated ?sense-for-sense? instea...