Search algorithms for statistical machine translation based on dynamic programming and pruning techniques

Ismael García-Varea, Francisco Casacuberta


Abstract
The increasing interest in the statistical approach to Machine Translation is due to the development of effective algorithms for training the probabilistic models proposed so far. However, one of the open problems with statistical machine translation is the design of efficient algorithms for translating a given input string. For some interesting models, only (good) approximate solutions can be found. Recently, a dynamic programming-like algorithm for the IBM-Model 2 has been proposed which is based on an iterative process of refinement solutions. A new dynamic programming-like algorithm is proposed here to deal with more complex IBM models (models 3 to 5). The computational cost of the algorithm is reduced by using an alignment-based pruning technique. Experimental results with the so-called “Tourist Task” are also presented.
Anthology ID:
2001.mtsummit-papers.22
Volume:
Proceedings of Machine Translation Summit VIII
Month:
September 18-22
Year:
2001
Address:
Santiago de Compostela, Spain
Editor:
Bente Maegaard
Venue:
MTSummit
SIG:
Publisher:
Note:
Pages:
Language:
URL:
https://aclanthology.org/2001.mtsummit-papers.22
DOI:
Bibkey:
Cite (ACL):
Ismael García-Varea and Francisco Casacuberta. 2001. Search algorithms for statistical machine translation based on dynamic programming and pruning techniques. In Proceedings of Machine Translation Summit VIII, Santiago de Compostela, Spain.
Cite (Informal):
Search algorithms for statistical machine translation based on dynamic programming and pruning techniques (García-Varea & Casacuberta, MTSummit 2001)
Copy Citation:
PDF:
https://aclanthology.org/2001.mtsummit-papers.22.pdf