logo Idiap Research Institute        
 [BibTeX] [Marc21]
Disambiguating Discourse Connectives for Statistical Machine Translation
Type of publication: Journal paper
Citation: Meyer_IEEETASLP_2015
Publication status: Published
Journal: IEEE/ACM Transactions on Audio, Speech and Language Processing
Volume: 23
Number: 7
Year: 2015
Month: July
Pages: 1184-1197
DOI: 10.1109/TASLP.2015.2422576
Abstract: This paper shows that the automatic labeling of discourse connectives with the relations they signal, prior to machine translation (MT), can be used by phrase-based statistical MT systems to improve their translations. This improvement is demonstrated here when translating from English to four target languages - French, German, Italian and Arabic - using several test sets from recent MT evaluation campaigns. Using automatically labeled data for training, tuning and testing MT systems is beneficial on condition that labels are sufficiently accurate, typically above 70%. To reach such an accuracy, a large array of features for discourse connective labeling (morpho-syntactic, semantic and discursive) are extracted using state-of-the-art tools and exploited in factored MT models. The translation of connectives is improved significantly, between 0.7% and 10% as measured with the dedicated ACT metric. The improvements depend mainly on the level of ambiguity of the connectives in the test sets.
Projects Idiap
Authors Meyer, Thomas
Hajlaoui, Najeh
Popescu-Belis, Andrei
Added by: [UNK]
Total mark: 0
  • Meyer_IEEETASLP_2015.pdf
       (Pre-print of IEEE/ACM TASLP paper.)