Neural Machine Translation with Noisy Lexical Constraints

Deng Cai; Guoping Huang; Huayang Li; Lemao Liu

REVIEW 4 major objections 4 minor 35 references

Neural Machine Translation with Noisy Lexical Constraints

T0 review · 4 major / 4 minor · reviewed 2026-08-14 · deepseek-v4-flash

Challenge this review Re-run · record.json Download PDF Read on arXiv ↗ Review JSON

Pith's one-line read The paper claims that a neural machine translation system can use user-supplied word constraints that may be wrong by treating them as external memories the decoder learns to use or ignore, rather than as hard requirements, and that this…

desk verdict A clean soft-memory framework for NMT with noisy lexical constraints, with consistent BLEU gains but a narrow noise simulation that limits external validity. read the letter →

arxiv 1908.04664 v4 pith:DBIDWENH submitted 2019-08-13 cs.CL

Huayang Li , Guoping Huang , Deng Cai , Lemao Liu This is my paper · ORCID

classification cs.CL

keywords noisylexicalconstraintsneuralmachinetranslationlexicallyconstraineddecodingexternalmemoryconstraintcorrectionsoftautomaticTransformer

verification ladder T0 review T1 audit T2 compute T3 formal

The pith

A machine-rendered reading of the paper's core claim, the machinery that carries it, and where it could break.

The reading

The paper tries to establish that noisy lexical constraints do not have to be obeyed: a neural machine translation decoder can be conditioned on learned representations of the constraint list, and then it will sometimes copy a constraint, sometimes ignore it, and sometimes replace it with something better. The authors show experimentally that this soft-memory treatment beats forced hard constrained decoding on Chinese-to-English and French-to-English translation when a fraction of the supplied words are wrong, at nearly the same decoding cost as an unconstrained Transformer. They also show that the same approach improves translation quality when constraints are generated automatically from the source text, with no user at all, even though those constraints are often incorrect. A sympathetic reader would take the contribution to be the removal of the perfect-constraint assumption that earlier constrained decoding depends on, plus a concrete mechanism for injecting imperfect side information.

What carries the argument

The mechanism is the constraint memory $E(c)$, a sequence of hidden vectors produced by a constraint encoder from the supplied word list, plus an integrator that feeds this memory into the decoder. The paper tests two encoders: a shallow one built from target word embeddings, which can mask out constraints already emitted, and a deep one reusing the Transformer encoder with shared target embeddings. It tests three integrators: a gated combination that interpolates the memory representation into the hidden state, a CopyNet-style constrained softmax that allocates probability mass to memory tokens, and a self-attention integrator that lets the decoder attend over the concatenation of its own history and the constraint memory. The common design choice doing the work is that inference is unconstrained beam search over $P(y|x,c;\theta)$, so the model can assign near-zero weight to a mistaken entry.

What would settle it

Take constraints from actual human users—typed terms, dictionary picks, or terminology edits—and run the same comparison: if the soft-memory models do not beat hard constrained decoding, or at least stay above the unconstrained Transformer, on that set, the claim that mistaken constraints can be corrected fails for realistic noise. A controlled variant would swap the translation-table neighbors for random target words and check whether the BLEU gains disappear, which would show the correction behavior depends on the specific noise distribution assumed in training.

Watch

Extended reading notes

Core claim

The central claim is that constraints should be external memories, not constraints: the model optimizes $P(y|x, c;\theta)$ with ordinary beam search, after a constraint encoder maps $c$ to memory vectors and an integrator blends them into the decoder. Because nothing forces the output to contain any of the supplied words, a wrong entry like "food" for a sentence about seeking assistance can be dropped, and a slightly off entry like "tourer" can be repaired to "tour". Across noise levels from one to five bad constraints, hard grid and dynamic beam decoding fall below the unconstrained baseline as noise grows, while the soft-memory models—especially the shallow encoder with gated combination or self-attention—remain stable or improve on both language pairs. With automatically generated constraints that are correct only about 38% of the time, the soft-memory models still beat the baseline by roughly 1.4 to 1.5 BLEU on Chinese-to-English and by about 1.3 BLEU on French-to-English in the best configuration, which the authors present as evidence that the framework has uses beyond user-provided hints.

Load-bearing premise

The load-bearing assumption is that replacing a correct constraint with a different but similar word from a word-to-word translation table, at a fixed 0.6 training noise rate, faithfully represents the mistakes real users would make; if real errors look different, the learned ability to correct them may not transfer.

Editorial extensions

If this is right

Constrained decoding systems that force constraints into the output should not be used with unreliable hints; under noise they can fall below the unconstrained baseline.
The soft-memory framework can deliver the benefit of lexical constraints at almost the decoding cost of a standard Transformer, instead of the added beam-search complexity of hard constrained decoding.
Automatically generated constraints from rare source words can improve translation without any human effort, extending the method to fully automatic workflows.
Because the model learns correction during training, a fixed noise-injection rate during fine-tuning is enough to produce robust behavior across a range of test noise rates.

Reading between the lines

Editorial extensions of the paper, not claims the author makes directly.

The gating signal learned by the model could plausibly be exposed as a confidence score for each supplied constraint, giving an interactive tool a way to flag hints the system distrusts; the paper does not report such an analysis.
Because the noise model is limited to translation-table neighbors, the strongest untested case is arbitrary user error—typos, homophones, or unrelated words—where the distributional similarity that helps correction would be absent.
The same "memory you may ignore" recipe transfers, in principle, to other imperfect side signals in generation, such as entity lists, style keywords, or retrieval-augmented context, whenever the signal is useful but not guaranteed correct.

Share X Bluesky LinkedIn Reddit HN

Signed reviews

No signed human review yet.

Editorial analysis

A structured set of objections, weighed in public.

Desk editor's note, referee report, and a circularity audit.

Desk Editor's Note

read the letter

Dear [Colleague],

The thing to know about this paper is that it's the first to look at lexical constraints in NMT when the constraints can be wrong. The authors don't try to make hard constrained decoding tolerate noise; they instead sidestep the problem by conditioning the decoder on a learned memory of the constraint list and letting the model decide how much to use each one. That's a clean idea, and the experiments back it up better than I expected.

The main positive is the consistency. Across Chinese-to-English and French-to-English, their soft-memory model (SE-ATTN) beats both GBS and DBA on perfect constraints, and it holds up much better than GBS when constraints are corrupted. The automatic-constraints scenario (Scenario 3) is a nice byproduct: they show that even when constraints come from a GIZA++ lookup table with no user input, the model still gains around 1.5 BLEU over the Transformer baseline. They also report runtime, which is close to the unconstrained model, so the approach is practically cheap.

Where it falls short: the noise model is narrow. Noisy constraints are always replacements sampled from a GIZA++ word-to-word table, and the automatic constraints use the same table. That means the model learns to handle one specific class of errors—target-side translation-table neighbors. Real user mistakes (typos, semantically related but not translation-table words, register errors) may not behave like that. The paper doesn't provide significance tests or error bars, and the BLEU differences are sometimes small (for example, DE-GATE vs Transformer in Table 4 is under a point). DBA is also absent from the noisy experiments, which is a gap. And the stress-test note is right: when all five constraints are noisy, SE-ATTN drops below the unconstrained Transformer (34.88 vs 36.22 on Ch-En), which suggests the model isn't truly 'correcting' bad constraints so much as partially ignoring them. The case study shows one example where it extracts a related meaning, but that's anecdotal.

None of this kills the paper. The central claim—that soft memory handling is more robust than hard constrained decoding under noisy hints—is supported. The limitations are mainly about the generality of the noise simulation and the strength of the evaluation, not about the architecture being incoherent.

Who's this for? Researchers working on interactive MT, constrained decoding, or human-in-the-loop systems. It's a solid empirical paper that deserves a proper peer review. A referee should ask for significance tests, a DBA comparison in the noisy setting, and ideally a more realistic noise source. I'd bring it to a reading group, though I wouldn't cite it heavily myself.

Recommendation: send it to review.

Referee Report

4 major / 4 minor

Summary. The paper proposes a soft-memory framework for lexically constrained neural machine translation that is intended to handle noisy user constraints. Instead of forcing constraints into the output as in grid beam search (GBS) or dynamic beam allocation (DBA), the framework encodes constraints into continuous memories and integrates them into the decoder through one of three integrators (gated combination, CopyNet, or a new self-attention mechanism), with either a shallow or deep constraint encoder. The authors evaluate three scenarios: perfect constraints extracted from references, simulated noisy constraints created by replacing reference words with GIZA++ translation-table neighbors, and automatically generated constraints from rare source words using the same translation table. Experiments on Chinese-to-English and French-to-English report BLEU gains over the unconstrained Transformer and over hard constrained decoding under simulated noise, and also report gains with automatically generated constraints. The paper claims to be the first to handle noisy lexical constraints and argues that a mistaken constraint can be corrected by the soft-memory treatment.

Significance. If the results hold, the framework is practically valuable because it removes the hard assumption that lexical constraints are perfect, it adds little decoding overhead relative to a standard Transformer, and it enables a new scenario where constraints are generated automatically without user involvement. The paper also provides a systematic comparison of six encoder/integrator combinations and a new self-attention integrator, which is a useful empirical contribution. The strengths are the two language pairs, the consistent BLEU tables, and the efficient decoding runtime. The main weakness is external validity: the noisy constraints in both the training and test stages come from the same GIZA++ replacement procedure, and the automatic-constraint scenario also relies on that table, so the learned correction behavior may be tuned to a narrow error class rather than to real user mistakes. In addition, the empirical claims would be stronger with significance tests or confidence intervals, because several reported differences are small, and the stronger DBA baseline is not evaluated in the noisy-constraint scenario.

major comments (4)

[Sections 4.2 and 4.3] The headline claim that a mistaken constraint can be corrected is supported only by a simulated noise process in which noisy constraints are GIZA++ translation-table neighbors of the correct word, and the automatic constraints in Section 4.3 are drawn from the same word-to-word table. Because the model is trained and tested on this single error class, the results do not establish robustness to realistic user errors such as typos, register errors, or semantically related words that are not translation-table neighbors. Please add a concrete test with human-annotated noisy constraints or with synthetic noise that is not drawn from the translation table, and report whether the framework still outperforms the unconstrained Transformer.
[Section 4.2] The training noise rate is fixed at 0.6 while test noise rates vary from 0.2 to 1.0. Without an ablation over training noise rates, the reported curves conflate the model's inherent robustness with a specific train/test noise distribution. Please report performance for models trained at several noise rates (for example 0.2, 0.4, 0.6, and 0.8) to show that the framework is robust rather than tuned to the single training value.
[Tables 3-5] BLEU differences are reported without significance tests, confidence intervals, or multiple-seed variance. On the French-to-English task several gains are under one BLEU point (for example Table 4, DE-GATE 67.85-67.98 versus TRANSFORMER 67.13, and Table 5, DE-GATE 67.92 versus 67.13), so the claimed substantial gains need statistical validation, such as bootstrap resampling of test sentences or paired significance tests.
[Section 5.3.2 and Tables 3-4] DBA, which is introduced and evaluated in Table 2 for perfect constraints, is not reported for the noisy-constraint scenario. Since DBA is a stronger hard-decoding baseline than GBS in some settings, its omission leaves open whether the claimed advantage over hard constrained decoding for noisy constraints generalizes beyond GBS. Please add DBA results to Tables 3 and 4 or justify its exclusion.

minor comments (4)

[Algorithm 1, line 4] The line says 'optimize θ according to Eq. (9)', but Eq. (9) is the inference argmax objective; the training objective is the negative log-likelihood defined in Section 3.2. Please correct this cross-reference.
[Section 3.1, Shallow Encoder] The text mentions that the shallow encoder can remove redundant constraints by a mask technique, but the masking procedure is not specified; please provide the masking details or pseudocode.
[Section 3.1, Eq. (8)] The notation in Eq. (8) is ambiguous: the left-hand side uses h^{d,l}_i while the right-hand side refers to a multi-head attention block whose arguments include \h^{d,l-1}_i and a concatenation, and the output of the final residual operation is not explicitly matched to h^{d,l}_i. Please align the notation with Eq. (3) and define all variables.
[Section 5.3.2, Table 3 discussion] The text states that all proposed models are more robust than GBS, which is true, but it does not note that SE-ATTN (34.88) falls below the unconstrained TRANSFORMER (36.22) when all five constraints are noisy; this tempers the correction claim and should be acknowledged.

Circularity Check

0 steps flagged · score 0.0 of 10

No significant circularity: the method is an empirical framework evaluated on held-out test sets, and the shared GIZA++ noise process is a validity caveat, not a definitional reduction.

full rationale

The paper's central claim is that soft memory-based constraint handling can tolerate and even 'correct' mistaken lexical constraints. This is supported by training the model on tuples of (source, reference, constraints) and evaluating it on held-out test sets with separate references, using beam search over the unconstrained objective in Eq. (9). Nothing in the derivation fits a parameter to the test set or renames a fitted constant as a prediction; the model could in principle fail, and in fact does degrade at high noise (Table 3: SE-ATTN reaches 34.88 BLEU with five noisy constraints versus 36.22 for the unconstrained Transformer). The shared GIZA++ word-to-word replacement procedure used for both training and test noise (Section 4.2) means the evaluation is confined to one simulated noise distribution, which is an external-validity limitation rather than a circularity: the test outputs are not entailed by the training objective, and the claimed gains are empirical facts about that simulation. Scenario 3 similarly trains and tests on automatically generated constraints from the same translation table, but the BLEU comparison on a held-out test set remains a genuine empirical outcome. There is no self-citation chain, uniqueness theorem, or ansatz smuggled in via citation that carries the argument. The paper is self-contained as an empirical study, so no circular step is exhibited.

Assumptions & free parameters 2 free parameters · 3 assumptions · 0 invented entities

This paper introduces no new physical entities, forces, or dimensions. The constraint memory is an architectural construct composed of known attention and memory mechanisms. The main implicit inputs are the hand-set experimental parameters (k=5, noise rate 0.6) and the assumption that the GIZA++ noise simulation matches real user errors.

free parameters (2)

Number of constraints per sentence k = 5 (experiments)
Used to generate constraints in all scenarios; while motivated by rare-word usefulness, the value is a hand-picked design choice rather than derived.
Training noise rate = 0.6
Scenario 2 trains with 60% noisy constraints while testing at 0.2 to 1.0; this mismatch is not analyzed and may influence robustness conclusions.

assumptions (3)

domain assumption User constraints are single words rather than phrases.
Stated in footnote 1 as 'without loss of generality'; restricts scope.
ad hoc to paper GIZA++ translation-table neighbors are reasonable replacements for noisy user constraints.
Section 4.2 simulates noise by replacing words with target words that share a source alignment; real human mistakes may not follow this distribution.
domain assumption The Transformer baseline and hard constrained decoding baselines (GBS and DBA) are correctly implemented.
The paper uses an in-house PyTorch Transformer (Section 5.2) and cites prior implementations; no code or checkpoints are provided.

how reviews work

0 comments

Cite this review

Pith. "Pith review of Neural Machine Translation with Noisy Lexical Constraints." pith.science (2026). https://pith.science/paper/DBIDWENH

@misc{pith2026190804664,
  author       = {Pith},
  title        = {Pith review of: Neural Machine Translation with Noisy Lexical Constraints},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/DBIDWENH}},
  note         = {Machine review of arXiv:1908.04664}
}

read the original abstract

Lexically constrained decoding for machine translation has shown to be beneficial in previous studies. Unfortunately, constraints provided by users may contain mistakes in real-world situations. It is still an open question that how to manipulate these noisy constraints in such practical scenarios. We present a novel framework that treats constraints as external memories. In this soft manner, a mistaken constraint can be corrected. Experiments demonstrate that our approach can achieve substantial BLEU gains in handling noisy constraints. These results motivate us to apply the proposed approach on a new scenario where constraints are generated without the help of users. Experiments show that our approach can indeed improve the translation quality with the automatically generated constraints.

Figures

Figures reproduced from arXiv: 1908.04664 by the authors.

**Figure 1.** The decoder of the proposed framework. an open question that how to improve the translation quality when constraints contain noises. To our knowledge, this paper makes the first attempt to incorporate NMT with noisy constraints. In order to make better use of noisy constraints, we propose a novel framework to correct the noises as shown in [PITH_FULL_IMAGE:figures/full_fig_p001_1.png] view at source ↗

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

35 extracted references · 20 canonical work pages

[1]

URL: " 'urlintro :=

ENTRY address author booktitle chapter edition editor howpublished institution journal key month note number organization pages publisher school series title type volume year eprint doi pubmed url lastchecked label extra.label sort.label short.list INTEGERS output.state before.all mid.sentence after.sentence after.block STRINGS urlintro eprinturl eprintpr...
[2]

write newline

" write newline "" before.all 'output.state := FUNCTION n.dashify 't := "" t empty not t #1 #1 substring "-" = t #1 #2 substring "--" = not "--" * t #2 global.max substring 't := t #1 #1 substring "-" = "-" * t #2 global.max substring 't := while if t #1 #1 substring * t #2 global.max substring 't := if while FUNCTION word.in bbl.in capitalize " " * FUNCT...
[4]

Dzmitry Bahdanau, Kyunghyun Cho, and Yoshua Bengio. 2014 b . Neural machine translation by jointly learning to align and translate. arXiv preprint arXiv:1409.0473

arXiv 2014
[5]

Ankur Bapna and Orhan Firat. 2019. Non-parametric adaptation for neural machine translation. arXiv preprint arXiv:1903.00058

work page Pith review arXiv 2019
[6]

Sergio Barrachina, Oliver Bender, Francisco Casacuberta, Jorge Civera, Elsa Cubel, Shahram Khadivi, Antonio Lagarda, Hermann Ney, Jes \'u s Tom \'a s, Enrique Vidal, et al. 2009. Statistical approaches to computer-assisted translation. Computational Linguistics, 35(1):3--28

work page 2009
[7]

Qian Cao and Deyi Xiong. 2018. Encoding gated translation memory into neural machine translation. In Proceedings of the 2018 Conference on Empirical Methods in Natural Language Processing, pages 3042--3047

work page 2018
[8]

Shanbo Cheng, Shujian Huang, Huadong Chen, Xin-Yu Dai, and Jiajun Chen. 2016. Primt: A pick-revise framework for interactive machine translation. In Proceedings of the 2016 Conference of the North American Chapter of the Association for Computational Linguistics: Human Language Technologies, pages 1240--1249

work page 2016
[9]

Georgiana Dinu, Prashant Mathur, Marcello Federico, and Yaser Al-Onaizan. 2019. Training neural machine translation to apply terminology constraints. arXiv preprint arXiv:1906.01105

arXiv 2019

Show all 35 references

[10]

Miguel Domingo, Alvaro Peris, and Francisco Casacuberta. 2016. Interactive-predictive translation based on multiple word-segments. In Proceedings of the 19th Annual Conference of the European Association for Machine Translation, pages 282--291

2016
[11]

Yang Feng, Shiyue Zhang, Andi Zhang, Dong Wang, and Andrew Abel. 2017. Memory-augmented neural machine translation. In Proceedings of the 2017 Conference on Empirical Methods in Natural Language Processing, pages 1390--1399, Copenhagen, Denmark. Association for Computational L...

2017
[12]

George Foster and Guy Lapalme. 2002. Text prediction for translators. Universit \'e de Montr \'e al

2002
[13]

Juri Ganitkevitch, Benjamin Van Durme, and Chris Callison-Burch. 2013. Ppdb: The paraphrase database. In Proceedings of the 2013 Conference of the North American Chapter of the Association for Computational Linguistics: Human Language Technologies, pages 758--764

2013
[14]

David Grangier and Michael Auli. 2018. Quickedit: Editing text & translations by crossing words out. In Proceedings of NAACL, pages 272--282

2018
[15]

Jiatao Gu, Zhengdong Lu, Hang Li, and Victor OK Li. 2016. Incorporating copying mechanism in sequence-to-sequence learning. arXiv preprint arXiv:1603.06393

2016 arXiv
[16]

Jiatao Gu, Yong Wang, Kyunghyun Cho, and Victor OK Li. 2018. Search engine guided non-parametric neural machine translation. In Proceedings of AAAI

2018
[17]

Eva Hasler, Adri \`a De Gispert, Gonzalo Iglesias, and Bill Byrne. 2018. Neural machine translation decoding with terminology constraints. In Proceedings of NAACL, pages 506--512

2018
[18]

Chris Hokamp and Qun Liu. 2017. Lexically constrained decoding for sequence generation using grid beam search. In Proceedings of the 55th Annual Meeting of the Association for Computational Linguistics (Volume 1: Long Papers), volume 1, pages 1535--1546

2017
[19]

ukasz Kaiser, Ofir Nachum, Aurko Roy, and Samy Bengio. 2017. Learning to remember rare events. arXiv preprint arXiv:1703.03129

2017 arXiv
[20]

Guillaume Klein, Yoon Kim, Yuntian Deng, Jean Senellart, and Alexander M Rush. 2017. Opennmt: Open-source toolkit for neural machine translation. arXiv preprint arXiv:1701.02810

2017 arXiv
[21]

Rebecca Knowles and Philipp Koehn. 2016. Neural interactive translation prediction. In Proceedings of the Association for Machine Translation in the Americas, pages 107--120

2016
[22]

Philipp Koehn. 2009. Statistical machine translation. Cambridge University Press

2009
[23]

Philipp Koehn, Hieu Hoang, Alexandra Birch, Chris Callison-Burch, Marcello Federico, Nicola Bertoldi, Brooke Cowan, Wade Shen, Christine Moran, Richard Zens, et al. 2007. Moses: Open source toolkit for statistical machine translation. In Proceedings of the 45th annual meeting ...

2007
[24]

Philipp Koehn, Franz Josef Och, and Daniel Marcu. 2003. Statistical phrase-based translation. In Proceedings of the 2003 Conference of the North American Chapter of the Association for Computational Linguistics on Human Language Technology-Volume 1, pages 48--54. Association f...

2003
[25]

Franz Josef Och and Hermann Ney. 2003. A systematic comparison of various statistical alignment models. Computational linguistics, 29(1):19--51

2003
[26]

Kishore Papineni, Salim Roukos, Todd Ward, and Wei-Jing Zhu. 2002. Bleu: a method for automatic evaluation of machine translation. In Proceedings of the 40th annual meeting on association for computational linguistics, pages 311--318. Association for Computational Linguistics

2002
[27]

\'A lvaro Peris, Miguel Domingo, and Francisco Casacuberta. 2017. Interactive neural machine translation. Computer Speech & Language, 45:201--220

2017
[28]

Matt Post and David Vilar. 2018. Fast lexically constrained decoding with dynamic beam allocation for neural machine translation. In Proceedings of NAACL

2018
[29]

Rico Sennrich, Barry Haddow, and Alexandra Birch. 2015. Neural machine translation of rare words with subword units. arXiv preprint arXiv:1508.07909

2015 arXiv
[30]

Ralf Steinberger, Bruno Pouliquen, Anna Widiger, Camelia Ignat, Tomaz Erjavec, Dan Tufis, and D \'a niel Varga. 2006. The jrc-acquis: A multilingual aligned parallel corpus with 20+ languages. arXiv preprint cs/0609058

2006 arXiv
[31]

Ilya Sutskever, Oriol Vinyals, and Quoc V Le. 2014. Sequence to sequence learning with neural networks. In Advances in neural information processing systems, pages 3104--3112

2014
[32]

Zhaopeng Tu, Yang Liu, Shuming Shi, and Tong Zhang. 2018. Learning to remember translation history with a continuous cache. Transactions of the Association of Computational Linguistics, 6:407--420

2018
[33]

Gomez, Lukasz Kaiser, and Illia Polosukhin

Ashish Vaswani, Noam Shazeer, Niki Parmar, Jakob Uszkoreit, Llion Jones, Aidan N. Gomez, Lukasz Kaiser, and Illia Polosukhin. 2017. Attention is all you need. CoRR

2017
[34]

Xing Wang, Zhengdong Lu, Zhaopeng Tu, Hang Li, Deyi Xiong, and Min Zhang. 2017. Neural machine translation advised by statistical machine translation. In AAAI, pages 3330--3336

2017
[35]

Joern Wuebker, Spence Green, John DeNero, Sasa Hasan, and Minh-Thang Luong. 2016. Models and inference for prefix-constrained machine translation. In Proceedings of the 54th Annual Meeting of the Association for Computational Linguistics (Volume 1: Long Papers), volume 1, pages 66--75

2016
[36]

Jiajun Zhang and Chengqing Zong. 2016. Bridging neural machine translation and bilingual dictionaries. arXiv preprint arXiv:1610.07272

2016 arXiv

Pith tools

Reviewed August 14, 2026 · model on record in the stance chip above.

[1] [1]

URL: " 'urlintro :=

ENTRY address author booktitle chapter edition editor howpublished institution journal key month note number organization pages publisher school series title type volume year eprint doi pubmed url lastchecked label extra.label sort.label short.list INTEGERS output.state before.all mid.sentence after.sentence after.block STRINGS urlintro eprinturl eprintpr...

[2] [2]

write newline

" write newline "" before.all 'output.state := FUNCTION n.dashify 't := "" t empty not t #1 #1 substring "-" = t #1 #2 substring "--" = not "--" * t #2 global.max substring 't := t #1 #1 substring "-" = "-" * t #2 global.max substring 't := while if t #1 #1 substring * t #2 global.max substring 't := if while FUNCTION word.in bbl.in capitalize " " * FUNCT...

[3] [4]

Dzmitry Bahdanau, Kyunghyun Cho, and Yoshua Bengio. 2014 b . Neural machine translation by jointly learning to align and translate. arXiv preprint arXiv:1409.0473

arXiv 2014

[4] [5]

Ankur Bapna and Orhan Firat. 2019. Non-parametric adaptation for neural machine translation. arXiv preprint arXiv:1903.00058

work page Pith review arXiv 2019

[5] [6]

Sergio Barrachina, Oliver Bender, Francisco Casacuberta, Jorge Civera, Elsa Cubel, Shahram Khadivi, Antonio Lagarda, Hermann Ney, Jes \'u s Tom \'a s, Enrique Vidal, et al. 2009. Statistical approaches to computer-assisted translation. Computational Linguistics, 35(1):3--28

work page 2009

[6] [7]

Qian Cao and Deyi Xiong. 2018. Encoding gated translation memory into neural machine translation. In Proceedings of the 2018 Conference on Empirical Methods in Natural Language Processing, pages 3042--3047

work page 2018

[7] [8]

Shanbo Cheng, Shujian Huang, Huadong Chen, Xin-Yu Dai, and Jiajun Chen. 2016. Primt: A pick-revise framework for interactive machine translation. In Proceedings of the 2016 Conference of the North American Chapter of the Association for Computational Linguistics: Human Language Technologies, pages 1240--1249

work page 2016

[8] [9]

Georgiana Dinu, Prashant Mathur, Marcello Federico, and Yaser Al-Onaizan. 2019. Training neural machine translation to apply terminology constraints. arXiv preprint arXiv:1906.01105

arXiv 2019