EMNLP 2021 论文列表

Dialogue models trained on human conversations inadvertently learn to generate toxic responses. In addition to producing explicitly offensive utterances, these models can also implicitly insult a group or individual by aligning themselves with an offensive statement. To better understand the dynamics of contextually offensive language, we investigate the stance of dialogue model responses in offensive Reddit conversations. Specifically, we create ToxiChat, a crowd-annotated dataset of 2,000 Reddit threads and model responses labeled with offensive language and stance. Our analysis reveals that 42% of human responses agree with toxic comments, whereas only 13% agree with safe comments. This undesirable behavior is learned by neural dialogue models, such as DialoGPT, which we show are two times more likely to agree with offensive comments. To enable automatic detection of offensive language, we fine-tuned transformer-based classifiers on ToxiChat that achieve 0.71 F1 for offensive labels and 0.53 Macro-F1 for stance labels. Finally, we quantify the effectiveness of controllable text generation (CTG) methods to mitigate the tendency of neural dialogue models to agree with offensive comments. Compared to the baseline, our best CTG model achieves a 19% reduction in agreement with offensive comments and produces 29% fewer offensive replies. Our work highlights the need for further efforts to characterize and analyze inappropriate behavior in dialogue models, in order to help make them safer.

A challenge in designing high-stakes language assessments is calibrating the test item difficulties, either a priori or from limited pilot test data. While prior work has addressed ‘cold start’ estimation of item difficulties without piloting, we devise a multi-task generalized linear model with BERT features to jump-start these estimates, rapidly improving their quality with as few as 500 test-takers and a small sample of item exposures (≈6 each) from a large item bank (≈4,000 items). Our joint model provides a principled way to compare test-taker proficiency, item difficulty, and language proficiency frameworks like the Common European Framework of Reference (CEFR). This also enables new item difficulty estimates without piloting them first, which in turn limits item exposure and thus enhances test item security. Finally, using operational data from the Duolingo English Test, a high-stakes English proficiency test, we find that the difficulty estimates derived using this method correlate strongly with lexico-grammatical features that correlate with reading complexity.

The task of news article image captioning aims to generate descriptive and informative captions for news article images. Unlike conventional image captions that simply describe the content of the image in general terms, news image captions follow journalistic guidelines and rely heavily on named entities to describe the image content, often drawing context from the whole article they are associated with. In this work, we propose a new approach to this task, motivated by caption guidelines that journalists follow. Our approach, Journalistic Guidelines Aware News Image Captioning (JoGANIC), leverages the structure of captions to improve the generation quality and guide our representation design. Experimental results, including detailed ablation studies, on two large-scale publicly available datasets show that JoGANIC substantially outperforms state-of-the-art methods both on caption generation and named entity related metrics.

We propose a novel task of jointly repairing program codes and generating commit messages. Code repair and commit message generation are two essential and related tasks for software development. However, existing work usually performs the two tasks independently. We construct a multilingual triple dataset including buggy code, fixed code, and commit messages for this novel task. We first introduce a cascaded method with two models, one is to generate the fixed code first, and the other generates the commit message based on the fixed and original codes. We enhance the cascaded method with different training approaches, including the teacher-student method, the multi-task method, and the back-translation method. To deal with the error propagation problem of the cascaded method, we also propose a joint model that can both repair the program code and generate the commit message in a unified framework. Massive experiments on our constructed buggy-fixed-commit dataset reflect the challenge of this task and that the enhanced cascaded model and the proposed joint model significantly outperform baselines in both quality of code and commit messages.

We study multi-answer retrieval, an under-explored problem that requires retrieving passages to cover multiple distinct answers for a given question. This task requires joint modeling of retrieved passages, as models should not repeatedly retrieve passages containing the same answer at the cost of missing a different valid answer. Prior work focusing on single-answer retrieval is limited as it cannot reason about the set of passages jointly. In this paper, we introduce JPR, a joint passage retrieval model focusing on reranking. To model the joint probability of the retrieved passages, JPR makes use of an autoregressive reranker that selects a sequence of passages, equipped with novel training and decoding algorithms. Compared to prior approaches, JPR achieves significantly better answer coverage on three multi-answer datasets. When combined with downstream question answering, the improved retrieval enables larger answer generation models since they need to consider fewer passages, establishing a new state-of-the-art.

Visual and textual modalities contribute complementary information about events described in multimedia documents. Videos contain rich dynamics and detailed unfoldings of events, while text describes more high-level and abstract concepts. However, existing event extraction methods either do not handle video or solely target video while ignoring other modalities. In contrast, we propose the first approach to jointly extract events from both video and text articles. We introduce the new task of Video MultiMedia Event Extraction and propose two novel components to build the first system towards this task. First, we propose the first self-supervised cross-modal event coreference model that can determine coreference between video events and text events without any manually annotated pairs. Second, we introduce the first cross-modal transformer architecture, which extracts structured event information from both videos and text documents. We also construct and will publicly release a new benchmark of video-article pairs, consisting of 860 video-article pairs with extensive annotations for evaluating methods on this task. Our experimental results demonstrate the effectiveness of our proposed method on our new benchmark dataset. We achieve 6.0% and 5.8% absolute F-score gain on multimodal event coreference resolution and multimedia event extraction.

Aspect terms extraction (ATE) and aspect sentiment classification (ASC) are two fundamental and fine-grained sub-tasks in aspect-level sentiment analysis (ALSA). In the textual analysis, joint extracting both aspect terms and sentiment polarities has been drawn much attention due to the better applications than individual sub-task. However, in the multi-modal scenario, the existing studies are limited to handle each sub-task independently, which fails to model the innate connection between the above two objectives and ignores the better applications. Therefore, in this paper, we are the first to jointly perform multi-modal ATE (MATE) and multi-modal ASC (MASC), and we propose a multi-modal joint learning approach with auxiliary cross-modal relation detection for multi-modal aspect-level sentiment analysis (MALSA). Specifically, we first build an auxiliary text-image relation detection module to control the proper exploitation of visual information. Second, we adopt the hierarchical framework to bridge the multi-modal connection between MATE and MASC, as well as separately visual guiding for each sub module. Finally, we can obtain all aspect-level sentiment polarities dependent on the jointly extracted specific aspects. Extensive experiments show the effectiveness of our approach against the joint textual approaches, pipeline and collapsed multi-modal approaches.

Parallel texts of Japanese and a non-pro-drop language have the potential of improving the performance of Japanese zero anaphora resolution (ZAR) because pronouns dropped in the former are usually mentioned explicitly in the latter. However, rule-based cross-lingual transfer is hampered by error propagation in an NLP pipeline and the frequent lack of transparency in translation correspondences. In this paper, we propose implicit transfer by injecting machine translation (MT) as an intermediate task between pretraining and ZAR. We employ a pretrained BERT model to initialize the encoder part of the encoder-decoder model for MT, and eject the encoder part for fine-tuning on ZAR. The proposed framework empirically demonstrates that ZAR performance can be improved by transfer learning from MT. In addition, we find that the incorporation of the masked language model training into MT leads to further gains.

Natural question generation (QG) aims to generate questions from a passage, and generated questions are answered from the passage. Most models with state-of-the-art performance model the previously generated text at each decoding step. However, (1) they ignore the rich structure information that is hidden in the previously generated text. (2) they ignore the impact of copied words on the passage. We perceive that information in previously generated words serves as auxiliary information in subsequent generation. To address these problems, we design the Iterative Graph Network-based Decoder (IGND) to model the previous generation using a Graph Neural Network at each decoding step. Moreover, our graph model captures dependency relations in the passage that boost the generation. Experimental results demonstrate that our model outperforms the state-of-the-art models with sentence-level QG tasks on SQuAD and MARCO datasets.

Most existing simultaneous machine translation (SiMT) systems are trained and evaluated on offline translation corpora. We argue that SiMT systems should be trained and tested on real interpretation data. To illustrate this argument, we propose an interpretation test set and conduct a realistic evaluation of SiMT trained on offline translations. Our results, on our test set along with 3 existing smaller scale language pairs, highlight the difference of up-to 13.83 BLEU score when SiMT models are evaluated on translation vs interpretation data. In the absence of interpretation training data, we propose a translation-to-interpretation (T2I) style transfer method which allows converting existing offline translations into interpretation-style data, leading to up-to 2.8 BLEU improvement. However, the evaluation gap remains notable, calling for constructing large-scale interpretation corpora better suited for evaluating and developing SiMT systems.

Automatic translation systems are known to struggle with rare words. Among these, named entities (NEs) and domain-specific terms are crucial, since errors in their translation can lead to severe meaning distortions. Despite their importance, previous speech translation (ST) studies have neglected them, also due to the dearth of publicly available resources tailored to their specific evaluation. To fill this gap, we i) present the first systematic analysis of the behavior of state-of-the-art ST systems in translating NEs and terminology, and ii) release NEuRoparl-ST, a novel benchmark built from European Parliament speeches annotated with NEs and terminology. Our experiments on the three language directions covered by our benchmark (en→es/fr/it) show that ST systems correctly translate 75–80% of terms and 65–70% of NEs, with very low performance (37–40%) on person names.

It is difficult to rank and evaluate the performance of grammatical error correction (GEC) systems, as a sentence can be rewritten in numerous correct ways. A number of GEC metrics have been used to evaluate proposed GEC systems; however, each system relies on either a comparison with one or more reference texts—in what is known as the gold standard for reference-based metrics—or a separate annotated dataset to fine-tune the reference-less metric. Reference-based systems have a low correlation with human judgement, cannot capture all the ways in which a sentence can be corrected, and require substantial work to develop a test dataset. We propose a reference-less GEC evaluation system that is strongly correlated with human judgement, solves the issues related to the use of a reference, and does not need another annotated dataset for fine-tuning. The proposed system relies solely on commonly available tools. Additionally, currently available reference-less metrics do not work properly when part of a sentence is repeated as opposed to reference-based metrics. In our proposed system, we look to address issues inherent in reference-less metrics and reference-based metrics.

Multi-hop reasoning has been widely studied in recent years to obtain more interpretable link prediction. However, we find in experiments that many paths given by these models are actually unreasonable, while little work has been done on interpretability evaluation for them. In this paper, we propose a unified framework to quantitatively evaluate the interpretability of multi-hop reasoning models so as to advance their development. In specific, we define three metrics, including path recall, local interpretability, and global interpretability for evaluation, and design an approximate strategy to calculate these metrics using the interpretability scores of rules. We manually annotate all possible rules and establish a benchmark. In experiments, we verify the effectiveness of our benchmark. Besides, we run nine representative baselines on our benchmark, and the experimental results show that the interpretability of current multi-hop reasoning models is less satisfactory and is 51.7% lower than the upper bound given by our benchmark. Moreover, the rule-based models outperform the multi-hop reasoning models in terms of performance and interpretability, which points to a direction for future research, i.e., how to better incorporate rule information into the multi-hop reasoning model. We will publish our codes and datasets upon acceptance.

The Uniform Information Density principle states that speakers plan their utterances to reduce fluctuations in the density of the information transmitted. In this paper, we test whether, and within which contextual units this principle holds in task-oriented dialogues. We show that there is evidence supporting the principle in written dialogues where participants play a cooperative reference game as well as in spoken dialogues involving instruction giving and following. Our study underlines the importance of identifying the relevant contextual components, showing that information content increases particularly within topically and referentially related contextual units.

The task of organizing a shuffled set of sentences into a coherent text has been used to evaluate a machine’s understanding of causal and temporal relations. We formulate the sentence ordering task as a conditional text-to-marker generation problem. We present Reorder-BART (Re-BART) that leverages a pre-trained Transformer-based model to identify a coherent order for a given set of shuffled sentences. The model takes a set of shuffled sentences with sentence-specific markers as input and generates a sequence of position markers of the sentences in the ordered text. Re-BART achieves the state-of-the-art performance across 7 datasets in Perfect Match Ratio (PMR) and Kendall’s tau. We perform evaluations in a zero-shot setting, showcasing that our model is able to generalize well across other datasets. We additionally perform several experiments to understand the functioning and limitations of our framework.

This paper investigates whether the power of the models pre-trained on text data, such as BERT, can be transferred to general token sequence classification applications. To verify pre-trained models’ transferability, we test the pre-trained models on text classification tasks with meanings of tokens mismatches, and real-world non-text token sequence classification data, including amino acid, DNA, and music. We find that even on non-text data, the models pre-trained on text converge faster, perform better than the randomly initialized models, and only slightly worse than the models using task-specific knowledge. We also find that the representations of the text and non-text pre-trained models share non-trivial similarities.

IrEne is an energy prediction system that accurately predicts the interpretable inference energy consumption of a wide range of Transformer-based NLP models. We present the IrEne-viz tool, an online platform for visualizing and exploring energy consumption of various Transformer-based models easily. Additionally, we release a public API that can be used to access granular information about energy consumption of transformer models and their components. The live demo is available at http://stonybrooknlp.github.io/irene/demo/.

Compared to fully manual translation, post-editing (PE) machine translation (MT) output can save time and reduce errors. Automatic word-level quality estimation (QE) aims to predict the correctness of words in MT output and holds great promise to aid PE by flagging problematic output. Quality of QE is crucial, as incorrect QE might lead to translators missing errors or wasting time on already correct MT output. Achieving accurate automatic word-level QE is very hard, and it is currently not known (i) at what quality threshold QE is actually beginning to be useful for human PE, and (ii), how to best present word-level QE information to translators. In particular, should word-level QE visualization indicate uncertainty of the QE model or not? In this paper, we address both research questions with real and simulated word-level QE, visualizations, and user studies, where time, subjective ratings, and quality of the final translations are assessed. Results show that current word-level QE models are not yet good enough to support PE. Instead, quality levels of > 80% F1 are required. For helpful quality levels, a visualization reflecting the uncertainty of the QE model is preferred. Our analysis further shows that speed gains achieved through QE are not merely a result of blindly trusting the QE system, but that the quality of the final translations also improves. The threshold results from the paper establish a quality goal for future word-level QE research.

Humans often employ figurative language use in communication, including during interactions with dialog systems. Thus, it is important for real-world dialog systems to be able to handle popular figurative language constructs like metaphor and simile. In this work, we analyze the performance of existing dialog models in situations where the input dialog context exhibits use of figurative language. We observe large gaps in handling of figurative language when evaluating the models on two open domain dialog datasets. When faced with dialog contexts consisting of figurative language, some models show very large drops in performance compared to contexts without figurative language. We encourage future research in dialog modeling to separately analyze and report results on figurative language in order to better test model capabilities relevant to real-world use. Finally, we propose lightweight solutions to help existing models become more robust to figurative language by simply using an external resource to translate figurative language to literal (non-figurative) forms while preserving the meaning to the best extent possible.