Transformer architecture achieves great success in abundant natural language processing tasks. The over-parameterization of the Transformer model has motivated plenty of works to alleviate its overfitting for superior performances. With some explorations, we find simple techniques such as dropout, can greatly boost model performance with a careful design. Therefore, in this paper, we integrate different dropout techniques into the training of Transformer models. Specifically, we propose an approach named UniDrop to unites three different dropout techniques from fine-grain to coarse-grain, i.e., feature dropout, structure dropout, and data dropout. Theoretically, we demonstrate that these three dropouts play different roles from regularization perspectives. Empirically, we conduct experiments on both neural machine translation and text classification benchmark datasets. Extensive results indicate that Transformer with UniDrop can achieve around 1.5 BLEU improvement on IWSLT14 translation tasks, and better accuracy for the classification even using strong pre-trained RoBERTa as backbone.

Non-autoregressive Transformer is a promising text generation model. However, current non-autoregressive models still fall behind their autoregressive counterparts in translation quality. We attribute this accuracy gap to the lack of dependency modeling among decoder inputs. In this paper, we propose CNAT, which learns implicitly categorical codes as latent variables into the non-autoregressive decoding. The interaction among these categorical codes remedies the missing dependencies and improves the model capacity. Experiment results show that our model achieves comparable or better performance in machine translation tasks than several strong baselines.

Subword segmentation algorithms have been a de facto choice when building neural machine translation systems. However, most of them need to learn a segmentation model based on some heuristics, which may produce sub-optimal segmentation. This can be problematic in some scenarios when the target language has rich morphological changes or there is not enough data for learning compact composition rules. Translating at fully character level has the potential to alleviate the issue, but empirical performances of character-based models has not been fully explored. In this paper, we present an in-depth comparison between character-based and subword-based NMT systems under three settings: translating to typologically diverse languages, training with low resource, and adapting to unseen domains. Experiment results show strong competitiveness of character-based models. Further analyses show that compared to subword-based models, character-based models are better at handling morphological phenomena, generating rare and unknown words, and more suitable for transferring to unseen domains.

Large-scale multi-label text classification (LMTC) tasks often face long-tailed label distributions, where many labels have few or even no training instances. Although current methods can exploit prior knowledge to handle these few/zero-shot labels, they neglect the meta-knowledge contained in the dataset that can guide models to learn with few samples. In this paper, for the first time, this problem is addressed from a meta-learning perspective. However, the simple extension of meta-learning approaches to multi-label classification is sub-optimal for LMTC tasks due to long-tailed label distribution and coexisting of few- and zero-shot scenarios. We propose a meta-learning approach named META-LMTC. Specifically, it constructs more faithful and more diverse tasks according to well-designed sampling strategies and directly incorporates the objective of adapting to new low-resource tasks into the meta-learning phase. Extensive experiments show that META-LMTC achieves state-of-the-art performance against strong baselines and can still enhance powerful BERTlike models.

This paper describes our proposed model for the Chinese Grammatical Error Diagnosis (CGED) task in NLPTEA2020. The goal of CGED is to use natural language processing techniques to automatically diagnose Chinese grammatical errors in sentences. To this end, we design and implement a CGED model named BERT with Score-feature Gates Error Diagnoser (BSGED), which is based on the BERT model, Bidirectional Long Short-Term Memory (BiLSTM) and conditional random field (CRF). In order to address the problem of losing partial-order relationships when embedding continuous feature items as with previous works, we propose a gating mechanism for integrating continuous feature items, which effectively retains the partial-order relationships between feature items. We perform LSTM processing on the encoding result of the BERT model, and further extract the sequence features. In the final test-set evaluation, we obtained the highest F1 score at the detection level and are among the top 3 F1 scores at the identification level.

Aspect-oriented Fine-grained Opinion Extraction (AFOE) aims at extracting aspect terms and opinion terms from review in the form of opinion pairs or additionally extracting sentiment polarity of aspect term to form opinion triplet. Because of containing several opinion factors, the complete AFOE task is usually divided into multiple subtasks and achieved in the pipeline. However, pipeline approaches easily suffer from error propagation and inconvenience in real-world scenarios. To this end, we propose a novel tagging scheme, Grid Tagging Scheme (GTS), to address the AFOE task in an end-to-end fashion only with one unified grid tagging task. Additionally, we design an effective inference strategy on GTS to exploit mutual indication between different opinion factors for more accurate extractions. To validate the feasibility and compatibility of GTS, we implement three different GTS models respectively based on CNN, BiLSTM, and BERT, and conduct experiments on the aspect-oriented opinion pair extraction and opinion triplet extraction datasets. Extensive experimental results indicate that GTS models outperform strong baselines significantly and achieve state-of-the-art performance.

Aspect-level sentiment classification (ASC) aims to detect the sentiment polarity of a given opinion target in a sentence. In neural network-based methods for ASC, most works employ the attention mechanism to capture the corresponding sentiment words of the opinion target, then aggregate them as evidence to infer the sentiment of the target. However, aspect-level datasets are all relatively small-scale due to the complexity of annotation. Data scarcity causes the attention mechanism sometimes to fail to focus on the corresponding sentiment words of the target, which finally weakens the performance of neural models. To address the issue, we propose a novel Attention Transfer Network (ATN) in this paper, which can successfully exploit attention knowledge from resource-rich document-level sentiment classification datasets to improve the attention capability of the aspect-level sentiment classification task. In the ATN model, we design two different methods to transfer attention knowledge and conduct experiments on two ASC benchmark datasets. Extensive experimental results show that our methods consistently outperform state-of-the-art works. Further analysis also validates the effectiveness of ATN.

Machine reading comprehension (MRC) is the task that asks a machine to answer questions based on a given context. For Chinese MRC, due to the non-literal and non-compositional semantic characteristics, Chinese idioms pose unique challenges for machines to understand. Previous studies tend to treat idioms separately without fully exploiting the relationship among them. In this paper, we first define the concept of literal meaning coverage to measure the consistency between semantics and literal meanings for Chinese idioms. With the definition, we prove that the literal meanings of many idioms are far from their semantics, and we also verify that the synonymic relationship can mitigate this inconsistency, which would be beneficial for idiom comprehension. Furthermore, to fully utilize the synonymic relationship, we propose the synonym knowledge enhanced reader. Specifically, for each idiom, we first construct a synonym graph according to the annotations from the high-quality synonym dictionary or the cosine similarity between the pre-trained idiom embeddings and then incorporate the graph attention network and gate mechanism to encode the graph. Experimental results on ChID, a large-scale Chinese idiom reading comprehension dataset, show that our model achieves state-of-the-art performance.

Recent proposed approaches have made promising progress in dialogue state tracking (DST). However, in multi-domain scenarios, ellipsis and reference are frequently adopted by users to express values that have been mentioned by slots from other domains. To handle these phenomena, we propose a Dialogue State Tracking with Slot Connections (DST-SC) model to explicitly consider slot correlations across different domains. Given a target slot, the slot connecting mechanism in DST-SC can infer its source slot and copy the source slot value directly, thus significantly reducing the difficulty of learning and reasoning. Experimental results verify the benefits of explicit slot connection modeling, and our model achieves state-of-the-art performance on MultiWOZ 2.0 and MultiWOZ 2.1 datasets.

Definition generation, which aims to automatically generate dictionary definitions for words, has recently been proposed to assist the construction of dictionaries and help people understand unfamiliar texts. However, previous works hardly consider explicitly modeling the “components” of definitions, leading to under-specific generation results. In this paper, we propose ESD, namely Explicit Semantic Decomposition for definition Generation, which explicitly decomposes the meaning of words into semantic components, and models them with discrete latent variables for definition generation. Experimental results show that achieves top results on WordNet and Oxford benchmarks, outperforming strong previous baselines.

Neural machine translation systems tend to fail on less decent inputs despite its significant efficacy, which may significantly harm the credibility of these systems—fathoming how and when neural-based systems fail in such cases is critical for industrial maintenance. Instead of collecting and analyzing bad cases using limited handcrafted error features, here we investigate this issue by generating adversarial examples via a new paradigm based on reinforcement learning. Our paradigm could expose pitfalls for a given performance metric, e.g., BLEU, and could target any given neural machine translation architecture. We conduct experiments of adversarial attacks on two mainstream neural machine translation architectures, RNN-search, and Transformer. The results show that our method efficiently produces stable attacks with meaning-preserving adversarial examples. We also present a qualitative and quantitative analysis for the preference pattern of the attack, demonstrating its capability of pitfall exposure.

Variational auto-encoders (VAEs) are widely used in natural language generation due to the regularization of the latent space. However, generating sentences from the continuous latent space does not explicitly model the syntactic information. In this paper, we propose to generate sentences from disentangled syntactic and semantic spaces. Our proposed method explicitly models syntactic information in the VAE’s latent space by using the linearized tree sequence, leading to better performance of language generation. Additionally, the advantage of sampling in the disentangled syntactic and semantic latent spaces enables us to perform novel applications, such as the unsupervised paraphrase generation and syntax transfer generation. Experimental results show that our proposed model achieves similar or better performance in various tasks, compared with state-of-the-art related work.

Current predominant neural machine translation (NMT) models often have a deep structure with large amounts of parameters, making these models hard to train and easily suffering from over-fitting. A common practice is to utilize a validation set to evaluate the training process and select the best checkpoint. Average and ensemble techniques on checkpoints can lead to further performance improvement. However, as these methods do not affect the training process, the system performance is restricted to the checkpoints generated in original training procedure. In contrast, we propose an online knowledge distillation method. Our method on-the-fly generates a teacher model from checkpoints, guiding the training process to obtain better performance. Experiments on several datasets and language pairs show steady improvement over a strong self-attention-based baseline system. We also provide analysis on data-limited setting against over-fitting. Furthermore, our method leads to an improvement in a machine reading experiment as well.

Opinion target extraction and opinion words extraction are two fundamental subtasks in Aspect Based Sentiment Analysis (ABSA). Recently, many methods have made progress on these two tasks. However, few works aim at extracting opinion targets and opinion words as pairs. In this paper, we propose a novel sequence labeling subtask for ABSA named TOWE (Target-oriented Opinion Words Extraction), which aims at extracting the corresponding opinion words for a given opinion target. A target-fused sequence labeling neural network model is designed to perform this task. The opinion target information is well encoded into context by an Inward-Outward LSTM. Then left and right contexts of the opinion target and the global context are combined to find the corresponding opinion words. We build four datasets for TOWE based on several popular ABSA benchmarks from laptop and restaurant reviews. The experimental results show that our proposed model outperforms the other compared methods significantly. We believe that our work may not only be helpful for downstream sentiment analysis task, but can also be used for pair-wise opinion summarization.

Distant supervision has obtained great progress on relation classification task. However, it still suffers from noisy labeling problem. Different from previous works that underutilize noisy data which inherently characterize the property of classification, in this paper, we propose RCEND, a novel framework to enhance Relation Classification by Exploiting Noisy Data. First, an instance discriminator with reinforcement learning is designed to split the noisy data into correctly labeled data and incorrectly labeled data. Second, we learn a robust relation classifier in semi-supervised learning way, whereby the correctly and incorrectly labeled data are treated as labeled and unlabeled data respectively. The experimental results show that our method outperforms the state-of-the-art models.

Previous studies have shown that neural machine translation (NMT) models can benefit from explicitly modeling translated () and untranslated () source contents as recurrent states (CITATION). However, this less interpretable recurrent process hinders its power to model the dynamic updating of and contents during decoding. In this paper, we propose to model the dynamic principles by explicitly separating source words into groups of translated and untranslated contents through parts-to-wholes assignment. The assignment is learned through a novel variant of routing-by-agreement mechanism (CITATION), namely Guided Dynamic Routing, where the translating status at each decoding step guides the routing process to assign each source word to its associated group (i.e., translated or untranslated content) represented by a capsule, enabling translation to be made from holistic context. Experiments show that our approach achieves substantial improvements over both Rnmt and Transformer by producing more adequate translations. Extensive analysis demonstrates that our method is highly interpretable, which is able to recognize the translated and untranslated contents as expected.

In sequence labeling, previous domain adaptation methods focus on the adaptation from the source domain to the entire target domain without considering the diversity of individual target domain samples, which may lead to negative transfer results for certain samples. Besides, an important characteristic of sequence labeling tasks is that different elements within a given sample may also have diverse domain relevance, which requires further consideration. To take the multi-level domain relevance discrepancy into account, in this paper, we propose a fine-grained knowledge fusion model with the domain relevance modeling scheme to control the balance between learning from the target domain data and learning from the source domain model. Experiments on three sequence labeling tasks show that our fine-grained knowledge fusion model outperforms strong baselines and other state-of-the-art sequence labeling domain adaptation methods.

Scenario-based question answering (SQA) has attracted increasing research attention. It typically requires retrieving and integrating knowledge from multiple sources, and applying general knowledge to a specific case described by a scenario. SQA widely exists in the medical, geography, and legal domains—both in practice and in the exams. In this paper, we introduce the GeoSQA dataset. It consists of 1,981 scenarios and 4,110 multiple-choice questions in the geography domain at high school level, where diagrams (e.g., maps, charts) have been manually annotated with natural language descriptions to benefit NLP research. Benchmark results on a variety of state-of-the-art methods for question answering, textual entailment, and reading comprehension demonstrate the unique challenges presented by SQA for future research.

Natural language sentences, being hierarchical, can be represented at different levels of granularity, like words, subwords, or characters. But most neural machine translation systems require the sentence to be represented as a sequence at a single level of granularity. It can be difficult to determine which granularity is better for a particular translation task. In this paper, we improve the model by incorporating multiple levels of granularity. Specifically, we propose (1) an encoder with character attention which augments the (sub)word-level representation with character-level information; (2) a decoder with multiple attentions that enable the representations from different levels of granularity to control the translation cooperatively. Experiments on three translation tasks demonstrate that our proposed models outperform the standard word-based model, the subword-based model, and a strong character-based model.

Existing neural machine translation systems do not explicitly model what has been translated and what has not during the decoding phase. To address this problem, we propose a novel mechanism that separates the source information into two parts: translated Past contents and untranslated Future contents, which are modeled by two additional recurrent layers. The Past and Future contents are fed to both the attention model and the decoder states, which provides Neural Machine Translation (NMT) systems with the knowledge of translated and untranslated contents. Experimental results show that the proposed approach significantly improves the performance in Chinese-English, German-English, and English-German translation tasks. Specifically, the proposed model outperforms the conventional coverage model in terms of both the translation quality and the alignment error rate.

In the encoder-decoder architecture for neural machine translation (NMT), the hidden states of the recurrent structures in the encoder and decoder carry the crucial information about the sentence. These vectors are generated by parameters which are updated by back-propagation of translation errors through time.We argue that propagating errors through the end-to-end recurrent structures are not a direct way of control the hidden vectors. In this paper, we propose to use word predictions as a mechanism for direct supervision. More specifically, we require these vectors to be able to predict the vocabulary in target sentence. Our simple mechanism ensures better representations in the encoder and decoder without using any extra data or annotation. It is also helpful in reducing the target side vocabulary and improving the decoding efficiency. Experiments on Chinese-English machine translation task show an average BLEU improvement by 4.53, respectively.

Neural parsers have benefited from automatically labeled data via dependency-context word embeddings. We investigate training character embeddings on a word-based context in a similar way, showing that the simple method improves state-of-the-art neural word segmentation models significantly, beating tri-training baselines for leveraging auto-segmented data.

Pairwise ranking methods are the most widely used discriminative training approaches for structure prediction problems in natural language processing (NLP). Decomposing the problem of ranking hypotheses into pairwise comparisons enables simple and efficient solutions. However, neglecting the global ordering of the hypothesis list may hinder learning. We propose a listwise learning framework for structure prediction problems such as machine translation. Our framework directly models the entire translation list’s ordering to learn parameters which may better fit the given listwise samples. Furthermore, we propose top-rank enhanced loss functions, which are more sensitive to ranking errors at higher positions. Experiments on a large-scale Chinese-English translation task show that both our listwise learning framework and top-rank enhanced listwise losses lead to significant improvements in translation quality.

Greedy transition-based parsers are appealing for their very fast speed, with reasonably high accuracies. In this paper, we build a fast shift-reduce neural constituent parser by using a neural network to make local decisions. One challenge to the parsing speed is the large hidden and output layer sizes caused by the number of constituent labels and branching options. We speed up the parser by using a hierarchical output layer, inspired by the hierarchical log-bilinear neural language model. In standard WSJ experiments, the neural parser achieves an almost 2.4 time speed up (320 sen/sec) compared to a non-hierarchical baseline without significant accuracy loss (89.06 vs 89.13 F-score).