This work uses a perceptron algorithm to learn more ro- bust weights to be used on out-of-domain corpora without the need for specialised data. For an Arabic-to ...
Robust Estimation of Feature Weights in Statistical Machine Translation ; Anthology ID: 2010.eamt-1.15 ; Volume: Proceedings of the 14th Annual Conference of the ...
The first stage, MERT, locates a good point in the weights space for the development set. 3. The second stage, Percpt, generalises the obtained values to be ...
This work proposes an improved English-Thai translation framework for non-timing aligned Parallel corpora using an improved alignment algorithm: Bleualign with ...
This work combines MERT with a perceptron training to obtain more robust weights. IN-DOMAIN TRAINING. An improvement of more than 2 points of BLEU with respect ...
Title, Robust Estimation of Feature Weights in Statistical Machine Translation. Publication Type, Conference Paper. Year of Publication, 2010.
Robust Estimation of Feature Weights in Statistical Machine Translation. lluis.marquez. Multilingual Online Translation.
People also ask
What is the use of robust estimation?
Can statistical techniques be used to perform the task of machine translation if so explain in brief?
What is syntax based statistical machine translation?
What is the difference between rule based machine translation and statistical machine translation?
Figure 2: Optimized feature weight values for each of the different tuning settings. Only a subset out of the 14 different tuning weights that are used by the ...
Weights of the various components in a standard Statistical Machine Translation model are usually estimated via Minimum Error Rate Training.
Clustering algorithms are introduced to enable robust estimation of parameters. The context-dependent alignment models are evaluated by comparison with gold ...