Difference between revisions of "An End-to-End Discriminative Approach to Machine Translation"
From Cohen Courses
Jump to navigationJump to search| Line 2: | Line 2: | ||
[http://dl.acm.org/ft_gateway.cfm?id=1220271&ftid=271584&dwn=1&CFID=137369265&CFTOKEN=55640909 paper] | [http://dl.acm.org/ft_gateway.cfm?id=1220271&ftid=271584&dwn=1&CFID=137369265&CFTOKEN=55640909 paper] | ||
| + | |||
| + | === Summary === | ||
| + | |||
| + | In this work, a discriminative approach to learn a translation model from parallel sentences. The translation task is viewed as the problem of finding the derivation h that mazimizes the translation score from the source s and target t. This score is calculated as a weighted feature combination, which this paper elaborates in detail. Parameter training is performed using a weighted perceptron algorithm. | ||
Revision as of 21:43, 5 November 2012
Citation
Summary
In this work, a discriminative approach to learn a translation model from parallel sentences. The translation task is viewed as the problem of finding the derivation h that mazimizes the translation score from the source s and target t. This score is calculated as a weighted feature combination, which this paper elaborates in detail. Parameter training is performed using a weighted perceptron algorithm.
