We focus on graph-to-sequence learning, which can be framed as transducing graph structures to sequences for text generation. To capture structural information associated with graphs, we investigate the problem of encoding graphs using graph convolutional networks (GCNs). Unlike various existing approaches where shallow architectures were used for capturing local structural information only, we introduce a dense connection strategy, proposing a novel Densely Connected Graph Convolutional Networks (DCGCNs). Such a deep architecture is able to integrate both local and non-local features to learn a better structural representation of a graph. Our model outperforms the state-of-the-art neural models significantly on AMRto-text generation and syntax-based neural machine translation.
This paper describes CAiRE's submission to the unsupervised machine translation track of the WMT'19 news shared task from German to Czech. We leverage a phrase-based statistical machine translation (PBSMT) model and a pre-trained language model to combine word-level neural machine translation (NMT) and subword-level NMT models without using any parallel data. We propose to solve the morphological richness problem of languages by training byte-pair encoding (BPE) embeddings for German and Czech separately, and they are aligned using MUSE (Conneau et al., 2018). To ensure the fluency and consistency of translations, a rescoring mechanism is proposed that reuses the pre-trained language model to select the translation candidates generated through beam search. Moreover, a series of pre-processing and post-processing approaches are applied to improve the quality of final translations.
Generating image descriptions in different languages is essential to satisfy users worldwide. However, it is prohibitively expensive to collect large-scale paired image-caption dataset for every target language which is critical for training descent image captioning models. Previous works tackle the unpaired cross-lingual image captioning problem through a pivot language, which is with the help of paired image-caption data in the pivot language and pivot-to-target machine translation models. However, such language-pivoted approach suffers from inaccuracy brought by the pivot-to-target translation, including disfluency and visual irrelevancy errors. In this paper, we propose to generate cross-lingual image captions with self-supervised rewards in the reinforcement learning framework to alleviate these two types of errors. We employ self-supervision from mono-lingual corpus in the target language to provide fluency reward, and propose a multi-level visual semantic matching model to provide both sentence-level and concept-level visual relevancy rewards. We conduct extensive experiments for unpaired cross-lingual image captioning in both English and Chinese respectively on two widely used image caption corpora. The proposed approach achieves significant performance improvement over state-of-the-art methods.
Back-translation is a widely used data augmentation technique which leverages target monolingual data. However, its effectiveness has been challenged since automatic metrics such as BLEU only show significant improvements for test examples where the source itself is a translation, or translationese. This is believed to be due to translationese inputs better matching the back-translated training data. In this work, we show that this conjecture is not empirically supported and that back-translation improves translation quality of both naturally occurring text as well as translationese according to professional human translators. We provide empirical evidence to support the view that back-translation is preferred by humans because it produces more fluent outputs. BLEU cannot capture human preferences because references are translationese when source sentences are natural text. We recommend complementing BLEU with a language model score to measure fluency.
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.
The learning rate warmup heuristic achieves remarkable success in stabilizing training, accelerating convergence and improving generalization for adaptive stochastic optimization algorithms like RMSprop and Adam. Here, we study its mechanism in details. Pursuing the theory behind warmup, we identify a problem of the adaptive learning rate (i.e., it has problematically large variance in the early stage), suggest warmup works as a variance reduction technique, and provide both empirical and theoretical evidence to verify our hypothesis. We further propose RAdam, a new variant of Adam, by introducing a term to rectify the variance of the adaptive learning rate. Extensive experimental results on image classification, language modeling, and neural machine translation verify our intuition and demonstrate the effectiveness and robustness of our proposed method. All implementations are available at: https://github.com/LiyuanLucasLiu/RAdam.
As the quality of machine translation rises and neural machine translation (NMT) is moving from sentence to document level translations, it is becoming increasingly difficult to evaluate the output of translation systems. We provide a test suite for WMT19 aimed at assessing discourse phenomena of MT systems participating in the News Translation Task. We have manually checked the outputs and identified types of translation errors that are relevant to document-level translation.
Natural language processing systems are often downstream of unreliable inputs: machine translation, optical character recognition, or speech recognition. For instance, virtual assistants can only answer your questions after understanding your speech. We investigate and mitigate the effects of noise from Automatic Speech Recognition systems on two factoid Question Answering (QA) tasks. Integrating confidences into the model and forced decoding of unknown words are empirically shown to improve the accuracy of downstream neural QA systems. We create and train models on a synthetic corpus of over 500,000 noisy sentences and evaluate on two human corpora from Quizbowl and Jeopardy! competitions.
Since deep learning became a key player in natural language processing (NLP), many deep learning models have been showing remarkable performances in a variety of NLP tasks, and in some cases, they are even outperforming humans. Such high performance can be explained by efficient knowledge representation of deep learning models. While many methods have been proposed to learn more efficient representation, knowledge distillation from pretrained deep networks suggest that we can use more information from the soft target probability to train other neural networks. In this paper, we propose a new knowledge distillation method self-knowledge distillation, based on the soft target probabilities of the training model itself, where multimode information is distilled from the word embedding space right below the softmax layer. Due to the time complexity, our method approximates the soft target probabilities. In experiments, we applied the proposed method to two different and fundamental NLP tasks: language model and neural machine translation. The experiment results show that our proposed method improves performance on the tasks.
One of the major problems in natural language processing (NLP) is the word sense disambiguation (WSD) problem. It is the task of computationally identifying the right sense of a polysemous word based on its context. Resolving the WSD problem boosts the accuracy of many NLP focused algorithms such as text classification and machine translation. In this paper, we introduce a new supervised algorithm for WSD, that is based on Kernel PCA and Semantic Diffusion Kernel, which is called Diffusion Kernel PCA (DKPCA). DKPCA grasps the semantic similarities within terms, and it is based on PCA. These properties enable us to perform feature extraction and dimension reduction guided by semantic similarities and within the algorithm. Our empirical results on SensEval data demonstrate that DKPCA achieves higher or very close accuracy results compared to SVM and KPCA with various well-known kernels when the labeled data ratio is meager. Considering the scarcity of labeled data, whereas large quantities of unlabeled textual data are easily accessible, these are highly encouraging first results to develop DKPCA further.
Machine Comprehension (MC) is one of the core problems in natural language processing, requiring both understanding of the natural language and knowledge about the world. Rapid progress has been made since the release of several benchmark datasets, and recently the state-of-the-art models even surpass human performance on the well-known SQuAD evaluation. In this paper, we transfer knowledge learned from machine comprehension to the sequence-to-sequence tasks to deepen the understanding of the text. We propose MacNet: a novel encoder-decoder supplementary architecture to the widely used attention-based sequence-to-sequence models. Experiments on neural machine translation (NMT) and abstractive text summarization show that our proposed framework can significantly improve the performance of the baseline models, and our method for the abstractive text summarization achieves the state-of-the-art results on the Gigaword dataset.
In the current work, we present a description of the system submitted to WMT 2019 News Translation Shared task. The system was created to translate news text from Lithuanian to English. To accomplish the given task, our system used a Word Embedding based Neural Machine Translation model to post edit the outputs generated by a Statistical Machine Translation model. The current paper documents the architecture of our model, descriptions of the various modules and the results produced using the same. Our system garnered a BLEU score of 17.6.
Although the problem of similar language translation has been an area of research interest for many years, yet it is still far from being solved. In this paper, we study the performance of two popular approaches: statistical and neural. We conclude that both methods yield similar results; however, the performance varies depending on the language pair. While the statistical approach outperforms the neural one by a difference of 6 BLEU points for the Spanish-Portuguese language pair, the proposed neural model surpasses the statistical one by a difference of 2 BLEU points for Czech-Polish. In the former case, the language similarity (based on perplexity) is much higher than in the latter case. Additionally, we report negative results for the system combination with back-translation. Our TALP-UPC system submission won 1st place for Czech-to-Polish and 2nd place for Spanish-to-Portuguese in the official evaluation of the 1st WMT Similar Language Translation task.
Neural machine translation (NMT) systems have been shown to give undesirable translation when a small change is made in the source sentence. In this paper, we study the behaviour of NMT systems when multiple changes are made to the source sentence. In particular, we ask the following question "Is it possible for an NMT system to predict same translation even when multiple words in the source sentence have been replaced?". To this end, we propose a soft-attention based technique to make the aforementioned word replacements. The experiments are conducted on two language pairs: English-German (en-de) and English-French (en-fr) and two state-of-the-art NMT systems: BLSTM-based encoder-decoder with attention and Transformer. The proposed soft-attention based technique achieves high success rate and outperforms existing methods like HotFlip by a significant margin for all the conducted experiments. The results demonstrate that state-of-the-art NMT systems are unable to capture the semantics of the source language. The proposed soft-attention based technique is an invariance-based adversarial attack on NMT systems. To better evaluate such attacks, we propose an alternate metric and argue its benefits in comparison with success rate.
We cast retrosynthesis as a machine translation problem by introducing a special Tensor2Tensor, an entire attention-based and fully data-driven model. Given a data set comprising about 50,000 diverse reactions extracted from USPTO patents, the model significantly outperforms seq2seq model (34.7%) on a top-1 accuracy by achieving 54.1%. For yielding better results, parameters such as batch size and training time are thoroughly investigated to train the model. Additionally, we offer a novel insight into the causes of grammatically invalid SMILES, and conduct a test in which experienced chemists pick out and analyze the "wrong" predictions that may be chemically plausible but differ from the ground truth. Actually, the effectiveness of our model is un-derestimated and the "true" top-1 accuracy can reach to 64.6%.
Dataflow visualization systems enable flexible visual data exploration by allowing the user to construct a dataflow diagram that composes query and visualization modules to specify system functionality. However learning dataflow diagram usage presents overhead that often discourages the user. In this work we design FlowSense, a natural language interface for dataflow visualization systems that utilizes state-of-the-art natural language processing techniques to assist dataflow diagram construction. FlowSense employs a semantic parser with special utterance tagging and special utterance placeholders to generalize to different datasets and dataflow diagrams. It explicitly presents recognized dataset and diagram special utterances to the user for dataflow context awareness. With FlowSense the user can expand and adjust dataflow diagrams more conveniently via plain English. We apply FlowSense to the VisFlow subset-flow visualization system to enhance its usability. We evaluate FlowSense by one case study with domain experts on a real-world data analysis problem and a formal user study.
In the current work, we present a description of the system submitted to WMT 2018 News Translation Shared task. The system was created to translate news text from Finnish to English. The system used a Character Based Neural Machine Translation model to accomplish the given task. The current paper documents the preprocessing steps, the description of the submitted system and the results produced using the same. Our system garnered a BLEU score of 12.9.
We introduce the metric using BERT (Bidirectional Encoder Representations from Transformers) (Devlin et al., 2019) for automatic machine translation evaluation. The experimental results of the WMT-2017 Metrics Shared Task dataset show that our metric achieves state-of-the-art performance in segment-level metrics task for all to-English language pairs.
In this paper we describe the CUNI translation system used for the unsupervised news shared task of the ACL 2019 Fourth Conference on Machine Translation (WMT19). We follow the strategy of Artexte et al. (2018b), creating a seed phrase-based system where the phrase table is initialized from cross-lingual embedding mappings trained on monolingual data, followed by a neural machine translation system trained on synthetic parallel data. The synthetic corpus was produced from a monolingual corpus by a tuned PBMT model refined through iterative back-translation. We further focus on the handling of named entities, i.e. the part of vocabulary where the cross-lingual embedding mapping suffers most. Our system reaches a BLEU score of 15.3 on the German-Czech WMT19 shared task.
We study an interesting problem in training neural network-based models for natural language generation tasks, which we call the \emph{representation degeneration problem}. We observe that when training a model for natural language generation tasks through likelihood maximization with the weight tying trick, especially with big training datasets, most of the learnt word embeddings tend to degenerate and be distributed into a narrow cone, which largely limits the representation power of word embeddings. We analyze the conditions and causes of this problem and propose a novel regularization method to address it. Experiments on language modeling and machine translation show that our method can largely mitigate the representation degeneration problem and achieve better performance than baseline algorithms.