Reinforcement Learning from Human Feedback (RLHF) and derivative techniques like Direct Preference Optimization (DPO) are task-alignment algorithms used to repurpose general, foundational models for specific tasks. We show that applying task-alignment to neural machine translation (NMT) addresses an existing task–data mismatch in NMT, leading to improvements across all languages of a multilingual model, even when task-alignment is only applied to a subset of those languages. We do so by introducing Direct Quality Optimization (DQO), a variant of DPO leveraging a pre-trained translation quality estimation model as a proxy for human preferences, and verify the improvements with both automatic metrics and through human evaluation.

With the internet growing increasingly multilingual, it is important to consider translating websites. However, professional translators are much more expensive than machines, and machine translation quality is continually increasing, so we must justify the cost of professional translation by measuring the effects of translation on website engagement, and how users interact with translations. This paper presents an in-the-wild study run on 2 websites fully translated into 15 and 11 languages respectively, where visitors with non-English preferred languages were randomized into being shown text translated by a professional translator, machine translated text, or untranslated English text. We find that both human and machine translations improve engagement, users rarely switch the page language manually, and that in-browser machine translation is often used when English is shown, particularly by users from countries with low English proficiency. We also release a dataset of interaction data collected during our studies, including 3,332,669 sessions from 190 countries across 2 websites.

While neural networks demonstrate a remarkable ability to model linguistic content, capturing contextual information related to a speaker’s conversational role is an open area of research. In this work, we analyze the effect of speaker role on language use through the game of Mafia, in which participants are assigned either an honest or a deceptive role. In addition to building a framework to collect a dataset of Mafia game records, we demonstrate that there are differences in the language produced by players with different roles. We confirm that classification models are able to rank deceptive players as more suspicious than honest ones based only on their use of language. Furthermore, we show that training models on two auxiliary tasks outperforms a standard BERT-based text classification approach. We also present methods for using our trained models to identify features that distinguish between player roles, which could be used to assist players during the Mafia game.

We introduce translation error correction (TEC), the task of automatically correcting human-generated translations. Imperfections in machine translations (MT) have long motivated systems for improving translations post-hoc with automatic post-editing. In contrast, little attention has been devoted to the problem of automatically correcting human translations, despite the intuition that humans make distinct errors that machines would be well-suited to assist with, from typos to inconsistencies in translation conventions. To investigate this, we build and release the Aced corpus with three TEC datasets (available at: github.com/lilt/tec). We show that human errors in TEC exhibit a more diverse range of errors and far fewer translation fluency errors than the MT errors in automatic post-editing datasets, suggesting the need for dedicated TEC models that are specialized to correct human errors. We show that pre-training instead on synthetic errors based on human errors improves TEC F-score by as much as 5.1 points. We conducted a human-in-the-loop user study with nine professional translation editors and found that the assistance of our TEC system led them to produce significantly higher quality revised translations.

We describe the task of bilingual markup transfer, which involves placing markup tags from a source sentence into a fixed target translation. This task arises in practice when a human translator generates the target translation without markup, and then the system infers the placement of markup tags. This task contrasts from previous work in which markup transfer is performed jointly with machine translation. We propose two novel metrics and evaluate several approaches based on unsupervised word alignments as well as a supervised neural sequence-to-sequence model. Our best approach achieves an average accuracy of 94.7% across six language pairs, indicating its potential usefulness for real-world localization tasks.

We present a set of assignments for a graduate-level NLP course. Assignments are designed to be interactive, easily gradable, and to give students hands-on experience with several key types of structure (sequences, tags, parse trees, and logical forms), modern neural architectures (LSTMs and Transformers), inference algorithms (dynamic programs and approximate search) and training methods (full and weak supervision). We designed assignments to build incrementally both within each assignment and across assignments, with the goal of enabling students to undertake graduate-level research in NLP by the end of the course.

Word alignment was once a core unsupervised learning task in natural language processing because of its essential role in training statistical machine translation (MT) models. Although unnecessary for training neural MT models, word alignment still plays an important role in interactive applications of neural machine translation, such as annotation transfer and lexicon injection. While statistical MT methods have been replaced by neural approaches with superior performance, the twenty-year-old GIZA++ toolkit remains a key component of state-of-the-art word alignment systems. Prior work on neural word alignment has only been able to outperform GIZA++ by using its output during training. We present the first end-to-end neural word alignment method that consistently outperforms GIZA++ on three data sets. Our approach repurposes a Transformer model trained for supervised translation to also serve as an unsupervised word alignment model in a manner that is tightly integrated and does not affect translation quality.

We propose an efficient batching strategy for variable-length decoding on GPU architectures. During decoding, when candidates terminate or are pruned according to heuristics, our streaming approach periodically “refills” the batch before proceeding with a selected subset of candidates. We apply our method to variable-width beam search on a state-of-the-art machine translation model. Our method decreases runtime by up to 71% compared to a fixed-width beam search baseline and 17% compared to a variable-width baseline, while matching baselines’ BLEU. Finally, experiments show that our method can speed up decoding in other domains, such as semantic and syntactic parsing.

Incremental domain adaptation, in which a system learns from the correct output for each input immediately after making its prediction for that input, can dramatically improve system performance for interactive machine translation. Users of interactive systems are sensitive to the speed of adaptation and how often a system repeats mistakes, despite being corrected. Adaptation is most commonly assessed using corpus-level BLEU- or TER-derived metrics that do not explicitly take adaptation speed into account. We find that these metrics often do not capture immediate adaptation effects, such as zero-shot and one-shot learning of domain-specific lexical items. To this end, we propose new metrics that directly evaluate immediate adaptation performance for machine translation. We use these metrics to choose the most suitable adaptation method from a range of different adaptation techniques for neural machine translation systems.

We propose and compare methods for gradient-based domain adaptation of self-attentive neural machine translation models. We demonstrate that a large proportion of model parameters can be frozen during adaptation with minimal or no reduction in translation quality by encouraging structured sparsity in the set of offset tensors during learning via group lasso regularization. We evaluate this technique for both batch and incremental adaptation across multiple data sets and language pairs. Our system architecture–combining a state-of-the-art self-attentive model with compact domain adaptation–provides high quality personalized machine translation that is both space and time efficient.