Melvin Johnson


2022

pdf
Multilingual Document-Level Translation Enables Zero-Shot Transfer From Sentences to Documents
Biao Zhang | Ankur Bapna | Melvin Johnson | Ali Dabirmoghaddam | Naveen Arivazhagan | Orhan Firat
Proceedings of the 60th Annual Meeting of the Association for Computational Linguistics (Volume 1: Long Papers)

Document-level neural machine translation (DocNMT) achieves coherent translations by incorporating cross-sentence context. However, for most language pairs there’s a shortage of parallel documents, although parallel sentences are readily available. In this paper, we study whether and how contextual modeling in DocNMT is transferable via multilingual modeling. We focus on the scenario of zero-shot transfer from teacher languages with document level data to student languages with no documents but sentence level data, and for the first time treat document-level translation as a transfer learning problem. Using simple concatenation-based DocNMT, we explore the effect of 3 factors on the transfer: the number of teacher languages with document level data, the balance between document and sentence level data at training, and the data condition of parallel documents (genuine vs. back-translated). Our experiments on Europarl-7 and IWSLT-10 show the feasibility of multilingual transfer for DocNMT, particularly on document-specific metrics. We observe that more teacher languages and adequate data balance both contribute to better transfer quality. Surprisingly, the transfer is less sensitive to the data condition, where multilingual DocNMT delivers decent performance with either back-translated or genuine document pairs.

pdf
DOCmT5: Document-Level Pretraining of Multilingual Language Models
Chia-Hsuan Lee | Aditya Siddhant | Viresh Ratnakar | Melvin Johnson
Findings of the Association for Computational Linguistics: NAACL 2022

In this paper, we introduce DOCmT5, a multilingual sequence-to-sequence language model pretrained with large-scale parallel documents. While previous approaches have focused on leveraging sentence-level parallel data, we try to build a general-purpose pretrained model that can understand and generate long documents. We propose a simple and effective pretraining objective - Document reordering Machine Translation (DrMT), in which the input documents that are shuffled and masked need to be translated. DrMT brings consistent improvements over strong baselines on a variety of document-level generation tasks, including over 12 BLEU points for seen-language pair document-level MT, over 7 BLEU points for unseen-language-pair document-level MT and over 3 ROUGE-1 points for seen-language pair cross-lingual summarization. We achieve state-of-the-art (SOTA) on WMT20 De-En and IWSLT15 Zh-En document translation tasks. We also conduct extensive analysis on various factors for document pretraining, including (1) the effects of pretraining data quality and (2) The effects of combining mono-lingual and cross-lingual pretraining. We plan to make our model checkpoints publicly available.

pdf
Domain Curricula for Code-Switched MT at MixMT 2022
Lekan Raheem | Maab Elrashid | Melvin Johnson | Julia Kreutzer
Proceedings of the Seventh Conference on Machine Translation (WMT)

In multilingual colloquial settings, it is a habitual occurrence to compose expressions of text or speech containing tokens or phrases of different languages, a phenomenon popularly known as code-switching or code-mixing (CMX). We present our approach and results for the Code-mixed Machine Translation (MixMT) shared task at WMT 2022: the task consists of two subtasks, monolingual to code-mixed machine translation (Subtask-1) and code-mixed to monolingual machine translation (Subtask-2). Most non-synthetic code-mixed data are from social media but gathering a significant amount of this kind of data would be laborious and this form of data has more writing variation than other domains, so for both subtasks, we experimented with data schedules for out-of-domain data. We jointly learn multiple domains of text by pretraining and fine-tuning, combined with a sentence alignment objective. We found that switching between domains caused improved performance in the domains seen earliest during training, but depleted the performance on the remaining domains. A continuous training run with strategically dispensed data of different domains showed a significantly improved performance over fine-tuning.

2021

pdf
HintedBT: Augmenting Back-Translation with Quality and Transliteration Hints
Sahana Ramnath | Melvin Johnson | Abhirut Gupta | Aravindan Raghuveer
Proceedings of the 2021 Conference on Empirical Methods in Natural Language Processing

Back-translation (BT) of target monolingual corpora is a widely used data augmentation strategy for neural machine translation (NMT), especially for low-resource language pairs. To improve effectiveness of the available BT data, we introduce HintedBT—a family of techniques which provides hints (through tags) to the encoder and decoder. First, we propose a novel method of using both high and low quality BT data by providing hints (as source tags on the encoder) to the model about the quality of each source-target pair. We don’t filter out low quality data but instead show that these hints enable the model to learn effectively from noisy data. Second, we address the problem of predicting whether a source token needs to be translated or transliterated to the target language, which is common in cross-script translation tasks (i.e., where source and target do not share the written script). For such cases, we propose training the model with additional hints (as target tags on the decoder) that provide information about the operation required on the source (translation or both translation and transliteration). We conduct experiments and detailed analyses on standard WMT benchmarks for three cross-script low/medium-resource language pairs: Hindi,Gujarati,Tamil-to-English. Our methods compare favorably with five strong and well established baselines. We show that using these hints, both separately and together, significantly improves translation quality and leads to state-of-the-art performance in all three language pairs in corresponding bilingual settings.

pdf
XTREME-R: Towards More Challenging and Nuanced Multilingual Evaluation
Sebastian Ruder | Noah Constant | Jan Botha | Aditya Siddhant | Orhan Firat | Jinlan Fu | Pengfei Liu | Junjie Hu | Dan Garrette | Graham Neubig | Melvin Johnson
Proceedings of the 2021 Conference on Empirical Methods in Natural Language Processing

Machine learning has brought striking advances in multilingual natural language processing capabilities over the past year. For example, the latest techniques have improved the state-of-the-art performance on the XTREME multilingual benchmark by more than 13 points. While a sizeable gap to human-level performance remains, improvements have been easier to achieve in some tasks than in others. This paper analyzes the current state of cross-lingual transfer learning and summarizes some lessons learned. In order to catalyze meaningful progress, we extend XTREME to XTREME-R, which consists of an improved set of ten natural language understanding tasks, including challenging language-agnostic retrieval tasks, and covers 50 typologically diverse languages. In addition, we provide a massively multilingual diagnostic suite and fine-grained multi-dataset evaluation capabilities through an interactive public leaderboard to gain a better understanding of such models.

pdf
Findings of the Shared Task on Machine Translation in Dravidian languages
Bharathi Raja Chakravarthi | Ruba Priyadharshini | Shubhanker Banerjee | Richard Saldanha | John P. McCrae | Anand Kumar M | Parameswari Krishnamurthy | Melvin Johnson
Proceedings of the First Workshop on Speech and Language Technologies for Dravidian Languages

This paper presents an overview of the shared task on machine translation of Dravidian languages. We presented the shared task results at the EACL 2021 workshop on Speech and Language Technologies for Dravidian Languages. This paper describes the datasets used, the methodology used for the evaluation of participants, and the experiments’ overall results. As a part of this shared task, we organized four sub-tasks corresponding to machine translation of the following language pairs: English to Tamil, English to Malayalam, English to Telugu and Tamil to Telugu which are available at https://competitions.codalab.org/competitions/27650. We provided the participants with training and development datasets to perform experiments, and the results were evaluated on unseen test data. In total, 46 research groups participated in the shared task and 7 experimental runs were submitted for evaluation. We used BLEU scores for assessment of the translations.

pdf
MergeDistill: Merging Language Models using Pre-trained Distillation
Simran Khanuja | Melvin Johnson | Partha Talukdar
Findings of the Association for Computational Linguistics: ACL-IJCNLP 2021

pdf
nmT5 - Is parallel data still relevant for pre-training massively multilingual language models?
Mihir Kale | Aditya Siddhant | Rami Al-Rfou | Linting Xue | Noah Constant | Melvin Johnson
Proceedings of the 59th Annual Meeting of the Association for Computational Linguistics and the 11th International Joint Conference on Natural Language Processing (Volume 2: Short Papers)

Recently, mT5 - a massively multilingual version of T5 - leveraged a unified text-to-text format to attain state-of-the-art results on a wide variety of multilingual NLP tasks. In this paper, we investigate the impact of incorporating parallel data into mT5 pre-training. We find that multi-tasking language modeling with objectives such as machine translation during pre-training is a straightforward way to improve performance on downstream multilingual and cross-lingual tasks. However, the gains start to diminish as the model capacity increases, suggesting that parallel data might not be as essential for larger models. At the same time, even at larger model sizes, we find that pre-training with parallel data still provides benefits in the limited labelled data regime

pdf
Explicit Alignment Objectives for Multilingual Bidirectional Encoders
Junjie Hu | Melvin Johnson | Orhan Firat | Aditya Siddhant | Graham Neubig
Proceedings of the 2021 Conference of the North American Chapter of the Association for Computational Linguistics: Human Language Technologies

Pre-trained cross-lingual encoders such as mBERT (Devlin et al., 2019) and XLM-R (Conneau et al., 2020) have proven impressively effective at enabling transfer-learning of NLP systems from high-resource languages to low-resource languages. This success comes despite the fact that there is no explicit objective to align the contextual embeddings of words/sentences with similar meanings across languages together in the same space. In this paper, we present a new method for learning multilingual encoders, AMBER (Aligned Multilingual Bidirectional EncodeR). AMBER is trained on additional parallel data using two explicit alignment objectives that align the multilingual representations at different granularities. We conduct experiments on zero-shot cross-lingual transfer learning for different tasks including sequence tagging, sentence retrieval and sentence classification. Experimental results on the tasks in the XTREME benchmark (Hu et al., 2020) show that AMBER obtains gains of up to 1.1 average F1 score on sequence tagging and up to 27.3 average accuracy on retrieval over the XLM-R-large model which has 3.2x the parameters of AMBER. Our code and models are available at http://github.com/junjiehu/amber.

2019

pdf
Massively Multilingual Neural Machine Translation
Roee Aharoni | Melvin Johnson | Orhan Firat
Proceedings of the 2019 Conference of the North American Chapter of the Association for Computational Linguistics: Human Language Technologies, Volume 1 (Long and Short Papers)

Multilingual Neural Machine Translation enables training a single model that supports translation from multiple source languages into multiple target languages. We perform extensive experiments in training massively multilingual NMT models, involving up to 103 distinct languages and 204 translation directions simultaneously. We explore different setups for training such models and analyze the trade-offs between translation quality and various modeling decisions. We report results on the publicly available TED talks multilingual corpus where we show that massively multilingual many-to-many models are effective in low resource settings, outperforming the previous state-of-the-art while supporting up to 59 languages in 116 translation directions in a single model. Our experiments on a large-scale dataset with 103 languages, 204 trained directions and up to one million examples per direction also show promising results, surpassing strong bilingual baselines and encouraging future work on massively multilingual NMT.

pdf
Small and Practical BERT Models for Sequence Labeling
Henry Tsai | Jason Riesa | Melvin Johnson | Naveen Arivazhagan | Xin Li | Amelia Archer
Proceedings of the 2019 Conference on Empirical Methods in Natural Language Processing and the 9th International Joint Conference on Natural Language Processing (EMNLP-IJCNLP)

We propose a practical scheme to train a single multilingual sequence labeling model that yields state of the art results and is small and fast enough to run on a single CPU. Starting from a public multilingual BERT checkpoint, our final model is 6x smaller and 27x faster, and has higher accuracy than a state-of-the-art multilingual baseline. We show that our model especially outperforms on low-resource languages, and works on codemixed input text without being explicitly trained on codemixed examples. We showcase the effectiveness of our method by reporting on part-of-speech tagging and morphological prediction on 70 treebanks and 48 languages.

2018

pdf
The Best of Both Worlds: Combining Recent Advances in Neural Machine Translation
Mia Xu Chen | Orhan Firat | Ankur Bapna | Melvin Johnson | Wolfgang Macherey | George Foster | Llion Jones | Mike Schuster | Noam Shazeer | Niki Parmar | Ashish Vaswani | Jakob Uszkoreit | Lukasz Kaiser | Zhifeng Chen | Yonghui Wu | Macduff Hughes
Proceedings of the 56th Annual Meeting of the Association for Computational Linguistics (Volume 1: Long Papers)

The past year has witnessed rapid advances in sequence-to-sequence (seq2seq) modeling for Machine Translation (MT). The classic RNN-based approaches to MT were first out-performed by the convolutional seq2seq model, which was then out-performed by the more recent Transformer model. Each of these new approaches consists of a fundamental architecture accompanied by a set of modeling and training techniques that are in principle applicable to other seq2seq architectures. In this paper, we tease apart the new architectures and their accompanying techniques in two ways. First, we identify several key modeling and training techniques, and apply them to the RNN architecture, yielding a new RNMT+ model that outperforms all of the three fundamental architectures on the benchmark WMT’14 English to French and English to German tasks. Second, we analyze the properties of each fundamental seq2seq architecture and devise new hybrid architectures intended to combine their strengths. Our hybrid models obtain further improvements, outperforming the RNMT+ model on both benchmark datasets.

2017

pdf
Google’s Multilingual Neural Machine Translation System: Enabling Zero-Shot Translation
Melvin Johnson | Mike Schuster | Quoc V. Le | Maxim Krikun | Yonghui Wu | Zhifeng Chen | Nikhil Thorat | Fernanda Viégas | Martin Wattenberg | Greg Corrado | Macduff Hughes | Jeffrey Dean
Transactions of the Association for Computational Linguistics, Volume 5

We propose a simple solution to use a single Neural Machine Translation (NMT) model to translate between multiple languages. Our solution requires no changes to the model architecture from a standard NMT system but instead introduces an artificial token at the beginning of the input sentence to specify the required target language. Using a shared wordpiece vocabulary, our approach enables Multilingual NMT systems using a single model. On the WMT’14 benchmarks, a single multilingual model achieves comparable performance for English→French and surpasses state-of-theart results for English→German. Similarly, a single multilingual model surpasses state-of-the-art results for French→English and German→English on WMT’14 and WMT’15 benchmarks, respectively. On production corpora, multilingual models of up to twelve language pairs allow for better translation of many individual pairs. Our models can also learn to perform implicit bridging between language pairs never seen explicitly during training, showing that transfer learning and zero-shot translation is possible for neural translation. Finally, we show analyses that hints at a universal interlingua representation in our models and also show some interesting examples when mixing languages.