Learning a Phrase-based Translation Model from Monolingual Data with Application to Domain Adaptation
Article Structure
Abstract
Currently, almost all of the statistical machine translation (SMT) models are trained with the parallel corpora in some specific domains.
Introduction
During the last decade, statistical machine translation has made great progress.
What Can We Learn with Phrase Pair Induction?
Readers may doubt that if phrase pair induction is performed only using bilingual lexicon and monolingual data, what new translation knowledge can be learned?
Probabilistic Bilingual Lexicon Acquisition
In order to induce the phrase pairs from the in-domain monolingual data for domain adaptation, the probabilistic bilingual lexicon is essential.
For each 1‘ E V[Sk] :
5: add IMap[l‘ ] into positionArray 6
Phrase Pair Refinement and Parameterization
5.1 Phrase Pair Refinement
Related Work
As far as we know, few researchers study phrase pair induction from only monolingual data.
Experiments
7.1 Experimental Setup
Conclusion and Future Work
This paper proposes a simple but effective method to induce phrase pairs from monolingual data.
Topics
phrase pair
Appears in 66 sentences as: Phrase Pair (3) Phrase pair (1) phrase pair (37) Phrase pairs (2) phrase pairs (29)
In Learning a Phrase-based Translation Model from Monolingual Data with Application to Domain Adaptation
- The main idea of our method is to divide the phrase-level translation rule induction into two steps: bilingual lexicon induction and phrase pair induction.Page 2, “Introduction”
- Since many researchers have studied the bilingual lexicon induction, in this paper, we mainly concentrate ourselves on phrase pair induction given a probabilistic bilingual lexicon and two in-domain large monolingual data (source and target language).Page 2, “Introduction”
- In addition, we will further introduce how to refine the induced phrase pairs and estimate the parameters of the induced phrase pairs , such as four standard translation features and phrase reordering feature used in the conventional phrase-based models (Koehn et al., 2007).Page 2, “Introduction”
- In the rest of this paper, we first explain with examples to show what new translation knowledge can be learned with our proposed phrase pair induction method (Section 2), and then we introduce the approach for probabilistic bilingual lexicon acquisition in Section 3.Page 2, “Introduction”
- In Section 4 and 5, we respectively present our method for phrase pair induction and introduce an approach for phrase pair refinement and parameter estimation.Page 2, “Introduction”
- Readers may doubt that if phrase pair induction is performed only using bilingual lexicon and monolingual data, what new translation knowledge can be learned?Page 2, “What Can We Learn with Phrase Pair Induction?”
- In contrast, phrase pair induction can make up for this deficiency to some extent.Page 2, “What Can We Learn with Phrase Pair Induction?”
- From the induced phrase pairs with our method, we have conducted a deep analysis and find that we can learn three kinds of new translation knowledge: 1) word reordering in a phrase pair ; 2) idioms; and 3) unknown word translations.Page 2, “What Can We Learn with Phrase Pair Induction?”
- The word order encoded in a phrase pair is difficult to learn in a word-based SMT.Page 2, “What Can We Learn with Phrase Pair Induction?”
- In the second example, the italic source word corresponds to two target words (in italic), and the phrase pair is an idiom which cannot be learned from word-based SMT.Page 2, “What Can We Learn with Phrase Pair Induction?”
- In the third example, as we learn from the source and target monolingual text that the words around the italic ones are translations with each other, thus we cannot only extract a new phrase pair but also learn a translation pair of unknown words in italic.Page 2, “What Can We Learn with Phrase Pair Induction?”
See all papers in Proc. ACL 2013 that mention phrase pair.
See all papers in Proc. ACL that mention phrase pair.
Back to top.
in-domain
Appears in 20 sentences as: in-domain (25)
In Learning a Phrase-based Translation Model from Monolingual Data with Application to Domain Adaptation
- Finally, they used the learned translation model directly to translate unseen data (Ravi and Knight, 2011; Nuhn et al., 2012) or incorporated the learned bilingual lexicon as a new in-domain translation resource into the phrase-based model which is trained with out-of-domain data to improve the domain adaptation performance in machine translation (Dou and Knight, 2012).Page 1, “Introduction”
- Since many researchers have studied the bilingual lexicon induction, in this paper, we mainly concentrate ourselves on phrase pair induction given a probabilistic bilingual lexicon and two in-domain large monolingual data (source and target language).Page 2, “Introduction”
- In order to induce the phrase pairs from the in-domain monolingual data for domain adaptation, the probabilistic bilingual lexicon is essential.Page 2, “Probabilistic Bilingual Lexicon Acquisition”
- In this paper, we acquire the probabilistic bilingual lexicon from two approaches: 1) build a bilingual lexicon from large-scale out-of-domain parallel data; 2) adopt a manually collected in-domain lexicon.Page 2, “Probabilistic Bilingual Lexicon Acquisition”
- This paper uses Chinese-to-English translation as a case study and electronic data is the in-domain data we focus on.Page 2, “Probabilistic Bilingual Lexicon Acquisition”
- In order to assign probabilities to each entry, we apply the Corpus Translation Probability which used in (Wu et al., 2008): given an in-domain source language monolingual data, we translate this data with the phrase-based model trained on the out-of-domain News data, the in-domain lexicon and the in-domain target language monolingual data (for language model estimation).Page 3, “Probabilistic Bilingual Lexicon Acquisition”
- One is using an in-domain probabilistic bilingual lexicon to extract sub-sentential parallel fragments from comparable corpora (Munteanu and Marcu, 2006; Quirk et al., 2007; Cettolo et al., 2010).Page 6, “Related Work”
- Munteanu and Marcu (2006) first extract the candidate parallel sentences from the comparable corpora and further extract the accurate sub-sentential bilingual fragments from the candidate parallel sentences using the in-domain probabilistic bilingual lexicon.Page 6, “Related Work”
- We have introduced the out-of-domain data and the electronic in-domain lexicon in Section 3.Page 6, “Experiments”
- Besides the in-domain lexicon, we have collected respectively 1 million monolingual sentences in electronic area from the web.Page 7, “Experiments”
- We construct two kinds of phrase-based models using Moses (Koehn et al., 2007): one uses out-of-domain data and the other uses in-domain data.Page 7, “Experiments”
See all papers in Proc. ACL 2013 that mention in-domain.
See all papers in Proc. ACL that mention in-domain.
Back to top.
translation probabilities
Appears in 13 sentences as: translation probabilities (7) Translation Probability (2) translation probability (5)
In Learning a Phrase-based Translation Model from Monolingual Data with Application to Domain Adaptation
- After using the log-likelihood-ratios algorithm2, we obtain a probabilistic bilingual lexicon with bidirectional translation probabilities from the out-of-domain data.Page 3, “Probabilistic Bilingual Lexicon Acquisition”
- It should be noted that there is no translation probability in this lexicon.Page 3, “Probabilistic Bilingual Lexicon Acquisition”
- In order to assign probabilities to each entry, we apply the Corpus Translation Probability which used in (Wu et al., 2008): given an in-domain source language monolingual data, we translate this data with the phrase-based model trained on the out-of-domain News data, the in-domain lexicon and the in-domain target language monolingual data (for language model estimation).Page 3, “Probabilistic Bilingual Lexicon Acquisition”
- With the source language data and its translation, we estimate the bidirectional translation probabilities for each entry in the original lexicon.Page 3, “Probabilistic Bilingual Lexicon Acquisition”
- For the entries whose translation probabilities are not estimated, we just assign a uniform probability.Page 3, “Probabilistic Bilingual Lexicon Acquisition”
- That is if a source word has n translations, then the translation probability of target word given the source word is 1/11.Page 3, “Probabilistic Bilingual Lexicon Acquisition”
- This refinement can be applied before finding the phrase pair with maximum probability (Line 12 in Figure 2) so that the duplicate words do not affect the calculation of translation probability of phrase pair.Page 5, “Phrase Pair Refinement and Parameterization”
- 5.2 Translation Probability EstimationPage 5, “Phrase Pair Refinement and Parameterization”
- It is well known that in the phrase-based SMT there are four translation probabilities and the reordering probability for each phrase pair.Page 5, “Phrase Pair Refinement and Parameterization”
- The translation probabilities in the traditional phrase-based SMT include bidirectional phrase translation probabilities and bidirectional lexical weights.Page 5, “Phrase Pair Refinement and Parameterization”
- In this paper, we borrow and extend the idea of (Klementiev et al., 2012) to calculate the phrase-level translation probability with context information in source and target monolingual corpus.Page 6, “Phrase Pair Refinement and Parameterization”
See all papers in Proc. ACL 2013 that mention translation probabilities.
See all papers in Proc. ACL that mention translation probabilities.
Back to top.
phrase-based
Appears in 11 sentences as: phrase-based (12)
In Learning a Phrase-based Translation Model from Monolingual Data with Application to Domain Adaptation
- In this paper, we take a step forward and propose a simple but effective method to induce a phrase-based model from the monolingual corpora given an au-tomatically-induced translation lexicon or a manually-edited translation dictionary.Page 1, “Abstract”
- Novel translation models, such as phrase-based models (Koehn et a., 2007), hierarchical phrase-based models (Chiang, 2007) and linguistically syntax-based models (Liu et a., 2006; Huang et al., 2006; Galley, 2006; Zhang et a1, 2008; Chiang, 2010; Zhang et al., 2011; Zhai et al., 2011, 2012) have been proposed and achieved higher and higher translation performance.Page 1, “Introduction”
- Finally, they used the learned translation model directly to translate unseen data (Ravi and Knight, 2011; Nuhn et al., 2012) or incorporated the learned bilingual lexicon as a new in-domain translation resource into the phrase-based model which is trained with out-of-domain data to improve the domain adaptation performance in machine translation (Dou and Knight, 2012).Page 1, “Introduction”
- level translation rules and learn a phrase-based model from the monolingual corpora.Page 2, “Introduction”
- In addition, we will further introduce how to refine the induced phrase pairs and estimate the parameters of the induced phrase pairs, such as four standard translation features and phrase reordering feature used in the conventional phrase-based models (Koehn et al., 2007).Page 2, “Introduction”
- The induced phrase-based model will be used to help domain adaptation for machine translation.Page 2, “Introduction”
- In order to assign probabilities to each entry, we apply the Corpus Translation Probability which used in (Wu et al., 2008): given an in-domain source language monolingual data, we translate this data with the phrase-based model trained on the out-of-domain News data, the in-domain lexicon and the in-domain target language monolingual data (for language model estimation).Page 3, “Probabilistic Bilingual Lexicon Acquisition”
- It is well known that in the phrase-based SMT there are four translation probabilities and the reordering probability for each phrase pair.Page 5, “Phrase Pair Refinement and Parameterization”
- The translation probabilities in the traditional phrase-based SMT include bidirectional phrase translation probabilities and bidirectional lexical weights.Page 5, “Phrase Pair Refinement and Parameterization”
- For the target-side monolingual data, they just use it to train language model, and for the source-side monolingual data, they employ a baseline (word-based SMT or phrase-based SMT trained with small-scale bitext) to first translate the source sentences, combining the source sentence and its target translation as a bilingual sentence pair, and then train a new phrase-base SMT with these pseudo sentence pairs.Page 6, “Related Work”
- We construct two kinds of phrase-based models using Moses (Koehn et al., 2007): one uses out-of-domain data and the other uses in-domain data.Page 7, “Experiments”
See all papers in Proc. ACL 2013 that mention phrase-based.
See all papers in Proc. ACL that mention phrase-based.
Back to top.
BLEU
Appears in 7 sentences as: BLEU (8)
In Learning a Phrase-based Translation Model from Monolingual Data with Application to Domain Adaptation
- We use BLEU (Papineni et al., 2002) score with shortest length penalty as the evaluation metric and apply the pairwise re-sampling approach (Koehn, 2004) to perform the significance test.Page 7, “Experiments”
- We can see from the table that the domain lexicon is much helpful and significantly outperforms the baseline with more than 4.0 BLEU points.Page 7, “Experiments”
- When it is enhanced with the in-domain language model, it can further improve the translation performance by more than 2.5 BLEU points.Page 7, “Experiments”
- From the results, we see that transductive learning can further improve the translation performance significantly by 0.6 BLEU points.Page 7, “Experiments”
- All the experiments are run based on the configuration with BLEU 13.41 in Table 4, and we call this configuration BestConfig.Page 7, “Experiments”
- When using the first 100k sentences for phrase pair induction, it obtains a significant improvement over the BestConfig by 0.65 BLEU points and can outperform the transductive learning method.Page 8, “Experiments”
- It outperforms the BestConfig significantly with an improvement of 1.42 BLEU points and it also performs much better than transductive learning method With gains of 0.82 BLEU points.Page 8, “Experiments”
See all papers in Proc. ACL 2013 that mention BLEU.
See all papers in Proc. ACL that mention BLEU.
Back to top.
domain adaptation
Appears in 7 sentences as: domain adaptation (7)
In Learning a Phrase-based Translation Model from Monolingual Data with Application to Domain Adaptation
- We apply our method for the domain adaptation task and the extensive experiments show that our proposed method can substantially improve the translation quality.Page 1, “Abstract”
- Finally, they used the learned translation model directly to translate unseen data (Ravi and Knight, 2011; Nuhn et al., 2012) or incorporated the learned bilingual lexicon as a new in-domain translation resource into the phrase-based model which is trained with out-of-domain data to improve the domain adaptation performance in machine translation (Dou and Knight, 2012).Page 1, “Introduction”
- The induced phrase-based model will be used to help domain adaptation for machine translation.Page 2, “Introduction”
- Section 6 will show the detailed experiments for the task of domain adaptation .Page 2, “Introduction”
- In order to induce the phrase pairs from the in-domain monolingual data for domain adaptation , the probabilistic bilingual lexicon is essential.Page 2, “Probabilistic Bilingual Lexicon Acquisition”
- Our purpose is to induce phrase pairs to improve translation quality for domain adaptation .Page 6, “Experiments”
- Extensive experiments on domain adaptation have shown that our method can significantly outperform previous methods which also focus on exploring the in-domain lexicon and monolingual data.Page 9, “Conclusion and Future Work”
See all papers in Proc. ACL 2013 that mention domain adaptation.
See all papers in Proc. ACL that mention domain adaptation.
Back to top.
language model
Appears in 7 sentences as: language model (8)
In Learning a Phrase-based Translation Model from Monolingual Data with Application to Domain Adaptation
- In order to assign probabilities to each entry, we apply the Corpus Translation Probability which used in (Wu et al., 2008): given an in-domain source language monolingual data, we translate this data with the phrase-based model trained on the out-of-domain News data, the in-domain lexicon and the in-domain target language monolingual data (for language model estimation).Page 3, “Probabilistic Bilingual Lexicon Acquisition”
- For the target-side monolingual data, they just use it to train language model , and for the source-side monolingual data, they employ a baseline (word-based SMT or phrase-based SMT trained with small-scale bitext) to first translate the source sentences, combining the source sentence and its target translation as a bilingual sentence pair, and then train a new phrase-base SMT with these pseudo sentence pairs.Page 6, “Related Work”
- For the out-of-domain data, we build the phrase table and reordering table using the 2.08 million Chinese-to-English sentence pairs, and we use the SRILM toolkit (Stolcke, 2002) to train the 5-gram English language model with the target part of the parallel sentences and the Xinhua portion of the English Gigaword.Page 7, “Experiments”
- An in-domain 5-gram English language model is trained with the target 1 million monolingual data.Page 7, “Experiments”
- (2008) regards the in-domain lexicon with corpus translation probability as another phrase table and further use the in-domain language model besides the out-of-domain language model .Page 7, “Experiments”
- When it is enhanced with the in-domain language model , it can further improve the translation performance by more than 2.5 BLEU points.Page 7, “Experiments”
- (2008) and Bertoldi and Federico (2009) employed the transductive learning to first translate the source-side monolingual data using the best configuration (baseline+in—domain lexicon+in-domain language model ) and obtain 1-best translation for each source-side sentence.Page 7, “Experiments”
See all papers in Proc. ACL 2013 that mention language model.
See all papers in Proc. ACL that mention language model.
Back to top.
machine translation
Appears in 6 sentences as: machine translation (6)
In Learning a Phrase-based Translation Model from Monolingual Data with Application to Domain Adaptation
- Currently, almost all of the statistical machine translation (SMT) models are trained with the parallel corpora in some specific domains.Page 1, “Abstract”
- During the last decade, statistical machine translation has made great progress.Page 1, “Introduction”
- Recently, more and more researchers concentrated on taking full advantage of the monolingual corpora in both source and target languages, and proposed methods for bilingual lexicon induction from nonparallel data (Rapp, 1995, 1999; Koehn and Knight, 2002; Haghighi et al., 2008; Daume III and J agarlamudi, 2011) and proposed unsupervised statistical machine translation (bilingual lexicon is a byproduct) with only monolingual corpora (Ravi and Knight, 2011; Nuhn et al., 2012; Dou and Knight, 2012).Page 1, “Introduction”
- The unsupervised statistical machine translation method (Ravi and Knight, 2011; Nuhn et al., 2012; Dou and Knight, 2012) viewed the translation task as a decipherment problem and designed a generative model with the objective function to maximize the likelihood of the source language monolingual data.Page 1, “Introduction”
- Finally, they used the learned translation model directly to translate unseen data (Ravi and Knight, 2011; Nuhn et al., 2012) or incorporated the learned bilingual lexicon as a new in-domain translation resource into the phrase-based model which is trained with out-of-domain data to improve the domain adaptation performance in machine translation (Dou and Knight, 2012).Page 1, “Introduction”
- The induced phrase-based model will be used to help domain adaptation for machine translation .Page 2, “Introduction”
See all papers in Proc. ACL 2013 that mention machine translation.
See all papers in Proc. ACL that mention machine translation.
Back to top.
BLEU points
Appears in 5 sentences as: BLEU points (6)
In Learning a Phrase-based Translation Model from Monolingual Data with Application to Domain Adaptation
- We can see from the table that the domain lexicon is much helpful and significantly outperforms the baseline with more than 4.0 BLEU points .Page 7, “Experiments”
- When it is enhanced with the in-domain language model, it can further improve the translation performance by more than 2.5 BLEU points .Page 7, “Experiments”
- From the results, we see that transductive learning can further improve the translation performance significantly by 0.6 BLEU points .Page 7, “Experiments”
- When using the first 100k sentences for phrase pair induction, it obtains a significant improvement over the BestConfig by 0.65 BLEU points and can outperform the transductive learning method.Page 8, “Experiments”
- It outperforms the BestConfig significantly with an improvement of 1.42 BLEU points and it also performs much better than transductive learning method With gains of 0.82 BLEU points .Page 8, “Experiments”
See all papers in Proc. ACL 2013 that mention BLEU points.
See all papers in Proc. ACL that mention BLEU points.
Back to top.
parallel sentences
Appears in 5 sentences as: parallel sentence (1) parallel sentences (5)
In Learning a Phrase-based Translation Model from Monolingual Data with Application to Domain Adaptation
- However, for the phrase-level probability, we cannot use maximum likelihood estimation since the phrase pairs are not extracted from parallel sentences .Page 5, “Phrase Pair Refinement and Parameterization”
- Munteanu and Marcu (2006) first extract the candidate parallel sentences from the comparable corpora and further extract the accurate sub-sentential bilingual fragments from the candidate parallel sentences using the in-domain probabilistic bilingual lexicon.Page 6, “Related Work”
- Thus, finding the candidate parallel sentences is not possible in our situation.Page 6, “Related Work”
- They are neither parallel nor comparable because we cannot even extract a small number of parallel sentence pairs from this monolingual data using the method of (Munteanu and Marcu, 2006).Page 7, “Experiments”
- For the out-of-domain data, we build the phrase table and reordering table using the 2.08 million Chinese-to-English sentence pairs, and we use the SRILM toolkit (Stolcke, 2002) to train the 5-gram English language model with the target part of the parallel sentences and the Xinhua portion of the English Gigaword.Page 7, “Experiments”
See all papers in Proc. ACL 2013 that mention parallel sentences.
See all papers in Proc. ACL that mention parallel sentences.
Back to top.
phrase table
Appears in 4 sentences as: phrase table (6)
In Learning a Phrase-based Translation Model from Monolingual Data with Application to Domain Adaptation
- For the out-of-domain data, we build the phrase table and reordering table using the 2.08 million Chinese-to-English sentence pairs, and we use the SRILM toolkit (Stolcke, 2002) to train the 5-gram English language model with the target part of the parallel sentences and the Xinhua portion of the English Gigaword.Page 7, “Experiments”
- For the in-domain electronic data, we first consider the lexicon as a phrase table in which we assign a constant 1.0 for each of the four probabilities, and then we combine this initial phrase table and the induced phrase pairs to form the new phrase table .Page 7, “Experiments”
- (2008) regards the in-domain lexicon with corpus translation probability as another phrase table and further use the in-domain language model besides the out-of-domain language model.Page 7, “Experiments”
- With the source-side sentences and their translations, the new phrase table and reordering table are built.Page 7, “Experiments”
See all papers in Proc. ACL 2013 that mention phrase table.
See all papers in Proc. ACL that mention phrase table.
Back to top.
sentence pairs
Appears in 4 sentences as: sentence pair (1) sentence pairs (4)
In Learning a Phrase-based Translation Model from Monolingual Data with Application to Domain Adaptation
- For each entry in LLR-lex, such as ([34], of), we can learn two kinds of information from the out-of-domain word-aligned sentence pairs : one is whether the target translation is before or after the translation of the preceding source-side word (Order); the other is whether the target translation is adjacent with the translation of the preceding source-side word (Adjacency).Page 5, “Phrase Pair Refinement and Parameterization”
- For the target-side monolingual data, they just use it to train language model, and for the source-side monolingual data, they employ a baseline (word-based SMT or phrase-based SMT trained with small-scale bitext) to first translate the source sentences, combining the source sentence and its target translation as a bilingual sentence pair, and then train a new phrase-base SMT with these pseudo sentence pairs .Page 6, “Related Work”
- They are neither parallel nor comparable because we cannot even extract a small number of parallel sentence pairs from this monolingual data using the method of (Munteanu and Marcu, 2006).Page 7, “Experiments”
- For the out-of-domain data, we build the phrase table and reordering table using the 2.08 million Chinese-to-English sentence pairs , and we use the SRILM toolkit (Stolcke, 2002) to train the 5-gram English language model with the target part of the parallel sentences and the Xinhua portion of the English Gigaword.Page 7, “Experiments”
See all papers in Proc. ACL 2013 that mention sentence pairs.
See all papers in Proc. ACL that mention sentence pairs.
Back to top.
statistical machine translation
Appears in 4 sentences as: statistical machine translation (4)
In Learning a Phrase-based Translation Model from Monolingual Data with Application to Domain Adaptation
- Currently, almost all of the statistical machine translation (SMT) models are trained with the parallel corpora in some specific domains.Page 1, “Abstract”
- During the last decade, statistical machine translation has made great progress.Page 1, “Introduction”
- Recently, more and more researchers concentrated on taking full advantage of the monolingual corpora in both source and target languages, and proposed methods for bilingual lexicon induction from nonparallel data (Rapp, 1995, 1999; Koehn and Knight, 2002; Haghighi et al., 2008; Daume III and J agarlamudi, 2011) and proposed unsupervised statistical machine translation (bilingual lexicon is a byproduct) with only monolingual corpora (Ravi and Knight, 2011; Nuhn et al., 2012; Dou and Knight, 2012).Page 1, “Introduction”
- The unsupervised statistical machine translation method (Ravi and Knight, 2011; Nuhn et al., 2012; Dou and Knight, 2012) viewed the translation task as a decipherment problem and designed a generative model with the objective function to maximize the likelihood of the source language monolingual data.Page 1, “Introduction”
See all papers in Proc. ACL 2013 that mention statistical machine translation.
See all papers in Proc. ACL that mention statistical machine translation.
Back to top.
translation model
Appears in 4 sentences as: translation model (2) translation models (2)
In Learning a Phrase-based Translation Model from Monolingual Data with Application to Domain Adaptation
- Recently, some research works study the unsupervised SMT for inducing a simple word-based translation model from the monolingual corpora.Page 1, “Abstract”
- Novel translation models , such as phrase-based models (Koehn et a., 2007), hierarchical phrase-based models (Chiang, 2007) and linguistically syntax-based models (Liu et a., 2006; Huang et al., 2006; Galley, 2006; Zhang et a1, 2008; Chiang, 2010; Zhang et al., 2011; Zhai et al., 2011, 2012) have been proposed and achieved higher and higher translation performance.Page 1, “Introduction”
- However, all of these state-of-the-art translation models rely on the parallel corpora to induce translation rules and estimate the corresponding parameters.Page 1, “Introduction”
- Finally, they used the learned translation model directly to translate unseen data (Ravi and Knight, 2011; Nuhn et al., 2012) or incorporated the learned bilingual lexicon as a new in-domain translation resource into the phrase-based model which is trained with out-of-domain data to improve the domain adaptation performance in machine translation (Dou and Knight, 2012).Page 1, “Introduction”
See all papers in Proc. ACL 2013 that mention translation model.
See all papers in Proc. ACL that mention translation model.
Back to top.
parallel corpora
Appears in 3 sentences as: parallel corpora (3)
In Learning a Phrase-based Translation Model from Monolingual Data with Application to Domain Adaptation
- Currently, almost all of the statistical machine translation (SMT) models are trained with the parallel corpora in some specific domains.Page 1, “Abstract”
- However, all of these state-of-the-art translation models rely on the parallel corpora to induce translation rules and estimate the corresponding parameters.Page 1, “Introduction”
- It is unfortunate that the parallel corpora are very expensive to collect and are usually not available for resource-poor languages and for many specific domains even in a resource-rich language pair.Page 1, “Introduction”
See all papers in Proc. ACL 2013 that mention parallel corpora.
See all papers in Proc. ACL that mention parallel corpora.
Back to top.
translation quality
Appears in 3 sentences as: translation quality (3)
In Learning a Phrase-based Translation Model from Monolingual Data with Application to Domain Adaptation
- We apply our method for the domain adaptation task and the extensive experiments show that our proposed method can substantially improve the translation quality .Page 1, “Abstract”
- Our purpose is to induce phrase pairs to improve translation quality for domain adaptation.Page 6, “Experiments”
- some good methods to explore the potential of the given data to improve the translation quality .Page 7, “Experiments”
See all papers in Proc. ACL 2013 that mention translation quality.
See all papers in Proc. ACL that mention translation quality.
Back to top.
word alignment
Appears in 3 sentences as: word alignment (2) words aligned (1)
In Learning a Phrase-based Translation Model from Monolingual Data with Application to Domain Adaptation
- We employ the same algorithm used in (Munteanu and Marcu, 2006) which first use the GIZA++ (with grow-diag-final-and heuristic) to obtain the word alignment between source and target words, and then calculate the association strength between the aligned words.Page 3, “Probabilistic Bilingual Lexicon Acquisition”
- If a target word in t is a gap word, we suppose there is a word alignment between the target gap word and the source-side null.Page 5, “For each 1‘ E V[Sk] :”
- According to our analysis, we find that the biggest problem is that in the target-side of the phrase pair, there are two or more identical words aligned to the same source-Page 5, “Phrase Pair Refinement and Parameterization”
See all papers in Proc. ACL 2013 that mention word alignment.
See all papers in Proc. ACL that mention word alignment.
Back to top.
