EMNLP 2021 论文列表

While pre-trained language models (PTLMs) have achieved noticeable success on many NLP tasks, they still struggle for tasks that require event temporal reasoning, which is essential for event-centric applications. We present a continual pre-training approach that equips PTLMs with targeted knowledge about event temporal relations. We design self-supervised learning objectives to recover masked-out event and temporal indicators and to discriminate sentences from their corrupted counterparts (where event or temporal indicators got replaced). By further pre-training a PTLM with these objectives jointly, we reinforce its attention to event and temporal information, yielding enhanced capability on event temporal reasoning. This **E**ffective **CON**tinual pre-training framework for **E**vent **T**emporal reasoning (ECONET) improves the PTLMs’ fine-tuning performances across five relation extraction and question answering tasks and achieves new or on-par state-of-the-art performances in most of our downstream tasks.

Generating informative and appropriate responses is challenging but important for building human-like dialogue systems. Although various knowledge-grounded conversation models have been proposed, these models have limitations in utilizing knowledge that infrequently occurs in the training data, not to mention integrating unseen knowledge into conversation generation. In this paper, we propose an Entity-Agnostic Representation Learning (EARL) method to introduce knowledge graphs to informative conversation generation. Unlike traditional approaches that parameterize the specific representation for each entity, EARL utilizes the context of conversations and the relational structure of knowledge graphs to learn the category representation for entities, which is generalized to incorporating unseen entities in knowledge graphs into conversation generation. Automatic and manual evaluations demonstrate that our model can generate more informative, coherent, and natural responses than baseline models.

Knowledge distillation (KD) has been proved effective for compressing large-scale pre-trained language models. However, existing methods conduct KD statically, e.g., the student model aligns its output distribution to that of a selected teacher model on the pre-defined training dataset. In this paper, we explore whether a dynamic knowledge distillation that empowers the student to adjust the learning procedure according to its competency, regarding the student performance and learning efficiency. We explore the dynamical adjustments on three aspects: teacher model adoption, data selection, and KD objective adaptation. Experimental results show that (1) proper selection of teacher model can boost the performance of student model; (2) conducting KD with 10% informative instances achieves comparable performance while greatly accelerates the training; (3) the student performance can be boosted by adjusting the supervision contribution of different alignment objective. We find dynamic knowledge distillation is promising and provide discussions on potential future directions towards more efficient KD methods.

Online conversations can sometimes take a turn for the worse, either due to systematic cultural differences, accidental misunderstandings, or mere malice. Automatically forecasting derailment in public online conversations provides an opportunity to take early action to moderate it. Previous work in this space is limited, and we extend it in several ways. We apply a pretrained language encoder to the task, which outperforms earlier approaches. We further experiment with shifting the training paradigm for the task from a static to a dynamic one to increase the forecast horizon. This approach shows mixed results: in a high-quality data setting, a longer average forecast horizon can be achieved at the cost of a small drop in F1; in a low-quality data setting, however, dynamic training propagates the noise and is highly detrimental to performance.

Incorporating lexical knowledge into deep learning models has been proved to be very effective for sequence labeling tasks. However, previous works commonly have difficulty dealing with large-scale dynamic lexicons which often cause excessive matching noise and problems of frequent updates. In this paper, we propose DyLex, a plug-in lexicon incorporation approach for BERT based sequence labeling tasks. Instead of leveraging embeddings of words in the lexicon as in conventional methods, we adopt word-agnostic tag embeddings to avoid re-training the representation while updating the lexicon. Moreover, we employ an effective supervised lexical knowledge denoising method to smooth out matching noise. Finally, we introduce a col-wise attention based knowledge fusion mechanism to guarantee the pluggability of the proposed framework. Experiments on ten datasets of three tasks show that the proposed framework achieves new SOTA, even with very large scale lexicons.

In this paper, we provide a bilingual parallel human-to-human recommendation dialog dataset (DuRecDial 2.0) to enable researchers to explore a challenging task of multilingual and cross-lingual conversational recommendation. The difference between DuRecDial 2.0 and existing conversational recommendation datasets is that the data item (Profile, Goal, Knowledge, Context, Response) in DuRecDial 2.0 is annotated in two languages, both English and Chinese, while other datasets are built with the setting of a single language. We collect 8.2k dialogs aligned across English and Chinese languages (16.5k dialogs and 255k utterances in total) that are annotated by crowdsourced workers with strict quality control procedure. We then build monolingual, multilingual, and cross-lingual conversational recommendation baselines on DuRecDial 2.0. Experiment results show that the use of additional English data can bring performance improvement for Chinese conversational recommendation, indicating the benefits of DuRecDial 2.0. Finally, this dataset provides a challenging testbed for future studies of monolingual, multilingual, and cross-lingual conversational recommendation.

Consistency Identification has obtained remarkable success on open-domain dialogue, which can be used for preventing inconsistent response generation. However, in contrast to the rapid development in open-domain dialogue, few efforts have been made to the task-oriented dialogue direction. In this paper, we argue that consistency problem is more urgent in task-oriented domain. To facilitate the research, we introduce CI-ToD, a novel dataset for Consistency Identification in Task-oriented Dialog system. In addition, we not only annotate the single label to enable the model to judge whether the system response is contradictory, but also provide more fine-grained labels (i.e., Dialogue History Inconsistency, User Query Inconsistency and Knowledge Base Inconsistency) to encourage model to know what inconsistent sources lead to it. Empirical results show that state-of-the-art methods only achieve 51.3%, which is far behind the human performance of 93.2%, indicating that there is ample room for improving consistency identification ability. Finally, we conduct exhaustive experiments and qualitative analysis to comprehend key challenges and provide guidance for future directions. All datasets and models are publicly available at https://github.com/yizhen20133868/CI-ToD.

Recently more attention has been given to adversarial attacks on neural networks for natural language processing (NLP). A central research topic has been the investigation of search algorithms and search constraints, accompanied by benchmark algorithms and tasks. We implement an algorithm inspired by zeroth order optimization-based attacks and compare with the benchmark results in the TextAttack framework. Surprisingly, we find that optimization-based methods do not yield any improvement in a constrained setup and slightly benefit from approximate gradient information only in unconstrained setups where search spaces are larger. In contrast, simple heuristics exploiting nearest neighbors without querying the target function yield substantial success rates in constrained setups, and nearly full success rate in unconstrained setups, at an order of magnitude fewer queries. We conclude from these results that current TextAttack benchmark tasks are too easy and constraints are too strict, preventing meaningful research on black-box adversarial text attacks.

User targeting is an essential task in the modern advertising industry: given a package of ads for a particular category of products (e.g., green tea), identify the online users to whom the ad package should be targeted. A (ad package specific) user targeting model is typically trained using historical clickthrough data: positive instances correspond to users who have clicked on an ad in the package before, whereas negative instances correspond to users who have not clicked on any ads in the package that were displayed to them. Collecting a sufficient amount of positive training data for training an accurate user targeting model, however, is by no means trivial. This paper focuses on the development of a method for automatic augmentation of the set of positive training instances. Experimental results on two datasets, including a real-world company dataset, demonstrate the effectiveness of our proposed method.

Despite constant improvements in machine translation quality, automatic poetry translation remains a challenging problem due to the lack of open-sourced parallel poetic corpora, and to the intrinsic complexities involved in preserving the semantics, style and figurative nature of poetry. We present an empirical investigation for poetry translation along several dimensions: 1) size and style of training data (poetic vs. non-poetic), including a zero-shot setup; 2) bilingual vs. multilingual learning; and 3) language-family-specific models vs. mixed-language-family models. To accomplish this, we contribute a parallel dataset of poetry translations for several language pairs. Our results show that multilingual fine-tuning on poetic text significantly outperforms multilingual fine-tuning on non-poetic text that is 35X larger in size, both in terms of automatic metrics (BLEU, BERTScore, COMET) and human evaluation metrics such as faithfulness (meaning and poetic style). Moreover, multilingual fine-tuning on poetic data outperforms bilingual fine-tuning on poetic data.

Existing techniques for mitigating dataset bias often leverage a biased model to identify biased instances. The role of these biased instances is then reduced during the training of the main model to enhance its robustness to out-of-distribution data. A common core assumption of these techniques is that the main model handles biased instances similarly to the biased model, in that it will resort to biases whenever available. In this paper, we show that this assumption does not hold in general. We carry out a critical investigation on two well-known datasets in the domain, MNLI and FEVER, along with two biased instance detection methods, partial-input and limited-capacity models. Our experiments show that in around a third to a half of instances, the biased model is unable to predict the main model’s behavior, highlighted by the significantly different parts of the input on which they base their decisions. Based on a manual validation, we also show that this estimate is highly in line with human interpretation. Our findings suggest that down-weighting of instances detected by bias detection methods, which is a widely-practiced procedure, is an unnecessary waste of training data. We release our code to facilitate reproducibility and future research.

We demonstrate a library for the integration of domain knowledge in deep learning architectures. Using this library, the structure of the data is expressed symbolically via graph declarations and the logical constraints over outputs or latent variables can be seamlessly added to the deep models. The domain knowledge can be defined explicitly, which improves the explainability of the models in addition to their performance and generalizability in the low-data regime. Several approaches for such integration of symbolic and sub-symbolic models have been introduced; however, there is no library to facilitate the programming for such integration in a generic way while various underlying algorithms can be used. Our library aims to simplify programming for such integration in both training and inference phases while separating the knowledge representation from learning algorithms. We showcase various NLP benchmark tasks and beyond. The framework is publicly available at Github(https://github.com/HLR/DomiKnowS).

Dialogue state tracking (DST), which estimates user goals given a dialogue context, is an essential component of task-oriented dialogue systems. Conventional DST models are usually trained offline, which requires a fixed dataset prepared in advance. This paradigm is often impractical in real-world applications since online dialogue systems usually involve continually emerging new data and domains. Therefore, this paper explores Domain-Lifelong Learning for Dialogue State Tracking (DLL-DST), which aims to continually train a DST model on new data to learn incessantly emerging new domains while avoiding catastrophically forgetting old learned domains. To this end, we propose a novel domain-lifelong learning method, called Knowledge Preservation Networks (KPN), which consists of multi-prototype enhanced retrospection and multi-strategy knowledge distillation, to solve the problems of expression diversity and combinatorial explosion in the DLL-DST task. Experimental results show that KPN effectively alleviates catastrophic forgetting and outperforms previous state-of-the-art lifelong learning methods by 4.25% and 8.27% of whole joint goal accuracy on the MultiWOZ benchmark and the SGD benchmark, respectively.

Linguistic representations derived from text alone have been criticized for their lack of grounding, i.e., connecting words to their meanings in the physical world. Vision-and- Language (VL) models, trained jointly on text and image or video data, have been offered as a response to such criticisms. However, while VL pretraining has shown success on multimodal tasks such as visual question answering, it is not yet known how the internal linguistic representations themselves compare to their text-only counterparts. This paper compares the semantic representations learned via VL vs. text-only pretraining for two recent VL models using a suite of analyses (clustering, probing, and performance on a commonsense question answering task) in a language-only setting. We find that the multimodal models fail to significantly outperform the text-only variants, suggesting that future work is required if multimodal pretraining is to be pursued as a means of improving NLP in general.

Online forums such as ChangeMyView have been explored to research aspects of persuasion and argumentative quality in language. While previous research has focused on arguments between a view-holder and a persuader, we explore the premise that apart from the merits of arguments, persuasion is influenced by the ambient social community. We hypothesize that comments from the rest of the community can either affirm the original view or implicitly exert pressure to change it. We develop a structured model to capture the ambient community’s sentiment towards the discussion and its effect on persuasion. Our experiments show that social features themselves are significantly predictive of persuasion (even without looking at the actual content of discussion), with performance comparable to some earlier approaches that use content features. Combining community and content features leads to overall performance of 78.5% on the persuasion prediction task. Our analyses suggest that the effect of social pressure is comparable to the difference between persuasive and non-persuasive language strategies in driving persuasion and that social pressure might be a causal factor for persuasion.

Many crowdsourced NLP datasets contain systematic artifacts that are identified only after data collection is complete. Earlier identification of these issues should make it easier to create high-quality training and evaluation data. We attempt this by evaluating protocols in which expert linguists work ‘in the loop’ during data collection to identify and address these issues by adjusting task instructions and incentives. Using natural language inference as a test case, we compare three data collection protocols: (i) a baseline protocol with no linguist involvement, (ii) a linguist-in-the-loop intervention with iteratively-updated constraints on the writing task, and (iii) an extension that adds direct interaction between linguists and crowdworkers via a chatroom. We find that linguist involvement does not lead to increased accuracy on out-of-domain test sets compared to baseline, and adding a chatroom has no effect on the data. Linguist involvement does, however, lead to more challenging evaluation data and higher accuracy on some challenge sets, demonstrating the benefits of integrating expert analysis during data collection.

Pretraining techniques leveraging enormous datasets have driven recent advances in text summarization. While folk explanations suggest that knowledge transfer accounts for pretraining’s benefits, little is known about why it works or what makes a pretraining task or dataset suitable. In this paper, we challenge the knowledge transfer story, showing that pretraining on documents consisting of character n-grams selected at random, we can nearly match the performance of models pretrained on real corpora. This work holds the promise of eliminating upstream corpora, which may alleviate some concerns over offensive language, bias, and copyright issues. To see whether the small residual benefit of using real data could be accounted for by the structure of the pretraining task, we design several tasks motivated by a qualitative study of summarization corpora. However, these tasks confer no appreciable benefit, leaving open the possibility of a small role for knowledge transfer.

Style is an integral part of natural language. However, evaluation methods for style measures are rare, often task-specific and usually do not control for content. We propose the modular, fine-grained and content-controlled similarity-based STyle EvaLuation framework (STEL) to test the performance of any model that can compare two sentences on style. We illustrate STEL with two general dimensions of style (formal/informal and simple/complex) as well as two specific characteristics of style (contrac’tion and numb3r substitution). We find that BERT-based methods outperform simple versions of commonly used style measures like 3-grams, punctuation frequency and LIWC-based approaches. We invite the addition of further tasks and task instances to STEL and hope to facilitate the improvement of style-sensitive measures.

People convey their intention and attitude through linguistic styles of the text that they write. In this study, we investigate lexicon usages across styles throughout two lenses: human perception and machine word importance, since words differ in the strength of the stylistic cues that they provide. To collect labels of human perception, we curate a new dataset, Hummingbird, on top of benchmarking style datasets. We have crowd workers highlight the representative words in the text that makes them think the text has the following styles: politeness, sentiment, offensiveness, and five emotion types. We then compare these human word labels with word importance derived from a popular fine-tuned style classifier like BERT. Our results show that the BERT often finds content words not relevant to the target style as important words used in style prediction, but humans do not perceive the same way even though for some styles (e.g., positive sentiment and joy) human- and machine-identified words share significant overlap for some styles.