Statistical Phrase-Based Translation.

Publication TypeConference Paper
Year of Publication2003
AuthorsKoehn, P.; Och, F.J.; Marcu, D.
Conference NameProceedings of the Human Language Technology Conference of the North American Chapter of the Association for Computational Linguistics (HLT-NAACL 2003)
Pagination127--133
Conference Start DateMay 27 - June 1
Conference LocationEdmonton, Canada
AbstractWe propose a new phrase-based translation model and decoding algorithm that enables us to evaluate and compare several, previously proposed phrase-based translation models. Within our framework, we carry out a large number of experiments to understand better and explain why phrase-based models outperform word-based models. Our empirical results, which hold for all examined language pairs, suggest that the highest levels of performance can be obtained through relatively simple means: heuristic learning of phrase translations from word-based alignments and lexical weighting of phrase translations. Surprisingly, learning phrases longer than three words and learning phrases from high-accuracywordlevel alignment models does not have a strong impact on performance. Learning only syntactically motivated phrases degrades the performance of our systems.
URLClick Here