EMNLP 2021 Papers

Writing mammography reports can be error-prone and time-consuming for radiologists. In this paper we propose a method to generate mammography reports given four images, corresponding to the four views used in screening mammography. To the best of our knowledge our work represents the first attempt to generate the mammography report using deep-learning. We propose an encoder-decoder model that includes an EfficientNet-based encoder and a Transformer-based decoder. We demonstrate that the Transformer-based attention mechanism can combine visual and semantic information to localize salient regions on the input mammograms and generate a visually interpretable report. The conducted experiments, including an evaluation by a certified radiologist, show the effectiveness of the proposed method.

To obtain high-quality sentence embeddings from pretrained language models (PLMs), they must either be augmented with additional pretraining objectives or finetuned on a large set of labeled text pairs. While the latter approach typically outperforms the former, it requires great human effort to generate suitable datasets of sufficient size. In this paper, we show how PLMs can be leveraged to obtain high-quality sentence embeddings without the need for labeled data, finetuning or modifications to the pretraining objective: We utilize the generative abilities of large and high-performing PLMs to generate entire datasets of labeled text pairs from scratch, which we then use for finetuning much smaller and more efficient models. Our fully unsupervised approach outperforms strong baselines on several semantic textual similarity datasets.

Math word problem (MWP) is a challenging and critical task in natural language processing. Many recent studies formalize MWP as a generation task and have adopted sequence-to-sequence models to transform problem descriptions to mathematical expressions. However, mathematical expressions are prone to minor mistakes while the generation objective does not explicitly handle such mistakes. To address this limitation, we devise a new ranking task for MWP and propose Generate & Rank, a multi-task framework based on a generative pre-trained language model. By joint training with generation and ranking, the model learns from its own mistakes and is able to distinguish between correct and incorrect expressions. Meanwhile, we perform tree-based disturbance specially designed for MWP and an online update to boost the ranker. We demonstrate the effectiveness of our proposed method on the benchmark and the results show that our method consistently outperforms baselines in all datasets. Particularly, in the classical Math23k, our method is 7% (78.4% to 85.4%) higher than the state-of-the-art. Code could be found at https://github.com/huawei-noah/noah-research.

Deep reinforcement learning provides a promising approach for text-based games in studying natural language communication between humans and artificial agents. However, the generalization still remains a big challenge as the agents depend critically on the complexity and variety of training tasks. In this paper, we address this problem by introducing a hierarchical framework built upon the knowledge graph-based RL agent. In the high level, a meta-policy is executed to decompose the whole game into a set of subtasks specified by textual goals, and select one of them based on the KG. Then a sub-policy in the low level is executed to conduct goal-conditioned reinforcement learning. We carry out experiments on games with various difficulty levels and show that the proposed method enjoys favorable generalizability.

This paper considers the unsupervised domain adaptation problem for neural machine translation (NMT), where we assume the access to only monolingual text in either the source or target language in the new domain. We propose a cross-lingual data selection method to extract in-domain sentences in the missing language side from a large generic monolingual corpus. Our proposed method trains an adaptive layer on top of multilingual BERT by contrastive learning to align the representation between the source and target language. This then enables the transferability of the domain classifier between the languages in a zero-shot manner. Once the in-domain data is detected by the classifier, the NMT model is then adapted to the new domain by jointly learning translation and domain discrimination tasks. We evaluate our cross-lingual data selection method on NMT across five diverse domains in three language pairs, as well as a real-world scenario of translation for COVID-19. The results show that our proposed method outperforms other selection baselines up to +1.5 BLEU score.

The lexical substitution task aims at generating a list of suitable replacements for a target word in context, ideally keeping the meaning of the modified text unchanged. While its usage has increased in recent years, the paucity of annotated data prevents the finetuning of neural models on the task, hindering the full fruition of recently introduced powerful architectures such as language models. Furthermore, lexical substitution is usually evaluated in a framework that is strictly bound to a limited vocabulary, making it impossible to credit appropriate, but out-of-vocabulary, substitutes. To assess these issues, we proposed GeneSis (Generating Substitutes in contexts), the first generative approach to lexical substitution. Thanks to a seq2seq model, we generate substitutes for a word according to the context it appears in, attaining state-of-the-art results on different benchmarks. Moreover, our approach allows silver data to be produced for further improving the performances of lexical substitution systems. Along with an extensive analysis of GeneSis results, we also present a human evaluation of the generated substitutes in order to assess their quality. We release the fine-tuned models, the generated datasets, and the code to reproduce the experiments at https://github.com/SapienzaNLP/genesis.

Robust sequence-to-sequence modelling is an essential task in the real world where the inputs are often noisy. Both user-generated and machine generated inputs contain various kinds of noises in the form of spelling mistakes, grammatical errors, character recognition errors, all of which impact downstream tasks and affect interpretability of texts. In this work, we devise a novel sequence-to-sequence architecture for detecting and correcting different real world and artificial noises (adversarial attacks) from English texts. Towards that we propose a modified Transformer-based encoder-decoder architecture that uses a gating mechanism to detect types of corrections required and accordingly corrects texts. Experimental results show that our gated architecture with pre-trained language models perform significantly better that the non-gated counterparts and other state-of-the-art error correction models in correcting spelling and grammatical errors. Extrinsic evaluation of our model on Machine Translation (MT) and Summarization tasks show the competitive performance of the model against other generative sequence-to-sequence models under noisy inputs.

Large-scale language models such as GPT-3 are excellent few-shot learners, allowing them to be controlled via natural text prompts. Recent studies report that prompt-based direct classification eliminates the need for fine-tuning but lacks data and inference scalability. This paper proposes a novel data augmentation technique that leverages large-scale language models to generate realistic text samples from a mixture of real samples. We also propose utilizing soft-labels predicted by the language models, effectively distilling knowledge from the large-scale language models and creating textual perturbations simultaneously. We perform data augmentation experiments on diverse classification tasks and show that our method hugely outperforms existing text augmentation methods. We also conduct experiments on our newly proposed benchmark to show that the augmentation effect is not only attributed to memorization. Further ablation studies and a qualitative analysis provide more insights into our approach.

Practical dialogue systems require robust methods of detecting out-of-scope (OOS) utterances to avoid conversational breakdowns and related failure modes. Directly training a model with labeled OOS examples yields reasonable performance, but obtaining such data is a resource-intensive process. To tackle this limited-data problem, previous methods focus on better modeling the distribution of in-scope (INS) examples. We introduce GOLD as an orthogonal technique that augments existing data to train better OOS detectors operating in low-data regimes. GOLD generates pseudo-labeled candidates using samples from an auxiliary dataset and keeps only the most beneficial candidates for training through a novel filtering mechanism. In experiments across three target benchmarks, the top GOLD model outperforms all existing methods on all key metrics, achieving relative gains of 52.4%, 48.9% and 50.3% against median baseline performance. We also analyze the unique properties of OOS data to identify key factors for optimally applying our proposed method.

Knowledge graphs are essential for numerous downstream natural language processing applications, but are typically incomplete with many facts missing. This results in research efforts on multi-hop reasoning task, which can be formulated as a search process and current models typically perform short distance reasoning. However, the long-distance reasoning is also vital with the ability to connect the superficially unrelated entities. To the best of our knowledge, there lacks a general framework that approaches multi-hop reasoning in mixed long-short distance reasoning scenarios. We argue that there are two key issues for a general multi-hop reasoning model: i) where to go, and ii) when to stop. Therefore, we propose a general model which resolves the issues with three modules: 1) the local-global knowledge module to estimate the possible paths, 2) the differentiated action dropout module to explore a diverse set of paths, and 3) the adaptive stopping search module to avoid over searching. The comprehensive results on three datasets demonstrate the superiority of our model with significant improvements against baselines in both short and long distance reasoning scenarios.

Targeted evaluations have found that machine translation systems often output incorrect gender in translations, even when the gender is clear from context. Furthermore, these incorrectly gendered translations have the potential to reflect or amplify social biases. We propose gender-filtered self-training (GFST) to improve gender translation accuracy on unambiguously gendered inputs. Our GFST approach uses a source monolingual corpus and an initial model to generate gender-specific pseudo-parallel corpora which are then filtered and added to the training data. We evaluate GFST on translation from English into five languages, finding that it improves gender accuracy without damaging generic quality. We also show the viability of GFST on several experimental settings, including re-training from scratch, fine-tuning, controlling the gender balance of the data, forward translation, and back-translation.

In this work, we propose a novel framework, Gradient Aligned Mutual Learning BERT (GAML-BERT), for improving the early exiting of BERT. GAML-BERT’s contributions are two-fold. We conduct a set of pilot experiments, which shows that mutual knowledge distillation between a shallow exit and a deep exit leads to better performances for both. From this observation, we use mutual learning to improve BERT’s early exiting performances, that is, we ask each exit of a multi-exit BERT to distill knowledge from each other. Second, we propose GA, a novel training method that aligns the gradients from knowledge distillation to cross-entropy losses. Extensive experiments are conducted on the GLUE benchmark, which shows that our GAML-BERT can significantly outperform the state-of-the-art (SOTA) BERT early exiting methods.

The availability of corpora has led to significant advances in training semantic parsers in English. Unfortunately, for languages other than English, annotated data is limited and so is the performance of the developed parsers. Recently, pretrained multilingual models have been proven useful for zero-shot cross-lingual transfer in many NLP tasks. What else does it require to apply a parser trained in English to other languages for zero-shot cross-lingual semantic parsing? Will simple language-independent features help? To this end, we experiment with six Discourse Representation Structure (DRS) semantic parsers in English, and generalize them to Italian, German and Dutch, where there are only a small number of manually annotated parses available. Extensive experiments show that despite its simplicity, adding Universal Dependency (UD) relations and Universal POS tags (UPOS) as model-agnostic features achieves surprisingly strong improvement on all parsers.

Masked language modeling (MLM), a self-supervised pretraining objective, is widely used in natural language processing for learning text representations. MLM trains a model to predict a random sample of input tokens that have been replaced by a [MASK] placeholder in a multi-class setting over the entire vocabulary. When pretraining, it is common to use alongside MLM other auxiliary objectives on the token or sequence level to improve downstream performance (e.g. next sentence prediction). However, no previous work so far has attempted in examining whether other simpler linguistically intuitive or not objectives can be used standalone as main pretraining objectives. In this paper, we explore five simple pretraining objectives based on token-level classification tasks as replacements of MLM. Empirical results on GLUE and SQUAD show that our proposed methods achieve comparable or better performance to MLM using a BERT-BASE architecture. We further validate our methods using smaller models, showing that pretraining a model with 41% of the BERT-BASE’s parameters, BERT-MEDIUM results in only a 1% drop in GLUE scores with our best objective.

In this paper, we introduce the task of predicting severity of age-restricted aspects of movie content based solely on the dialogue script. We first investigate categorizing the ordinal severity of movies on 5 aspects: Sex, Violence, Profanity, Substance consumption, and Frightening scenes. The problem is handled using a siamese network-based multitask framework which concurrently improves the interpretability of the predictions. The experimental results show that our method outperforms the previous state-of-the-art model and provides useful information to interpret model predictions. The proposed dataset and source code are publicly available at our GitHub repository.

Cross-lingual entity alignment (EA) aims to find the equivalent entities between crosslingual KGs (Knowledge Graphs), which is a crucial step for integrating KGs. Recently, many GNN-based EA methods are proposed and show decent performance improvements on several public datasets. However, existing GNN-based EA methods inevitably inherit poor interpretability and low efficiency from neural networks. Motivated by the isomorphic assumption of GNN-based methods, we successfully transform the cross-lingual EA problem into an assignment problem. Based on this re-definition, we propose a frustratingly Simple but Effective Unsupervised entity alignment method (SEU) without neural networks. Extensive experiments have been conducted to show that our proposed unsupervised approach even beats advanced supervised methods across all public datasets while having high efficiency, interpretability, and stability.

Pre-trained language models perform well on a variety of linguistic tasks that require symbolic reasoning, raising the question of whether such models implicitly represent abstract symbols and rules. We investigate this question using the case study of BERT’s performance on English subject–verb agreement. Unlike prior work, we train multiple instances of BERT from scratch, allowing us to perform a series of controlled interventions at pre-training time. We show that BERT often generalizes well to subject–verb pairs that never occurred in training, suggesting a degree of rule-governed behavior. We also find, however, that performance is heavily influenced by word frequency, with experiments showing that both the absolute frequency of a verb form, as well as the frequency relative to the alternate inflection, are causally implicated in the predictions BERT makes at inference time. Closer analysis of these frequency effects reveals that BERT’s behavior is consistent with a system that correctly applies the SVA rule in general but struggles to overcome strong training priors and to estimate agreement features (singular vs. plural) on infrequent lexical items.

Sentence-level extractive text summarization aims to select important sentences from a given document. However, it is very challenging to model the importance of sentences. In this paper, we propose a novel Frame Semantic-Enhanced Sentence Modeling for Extractive Summarization, which leverages Frame semantics to model sentences from both intra-sentence level and inter-sentence level, facilitating the text summarization task. In particular, intra-sentence level semantics leverage Frames and Frame Elements to model internal semantic structure within a sentence, while inter-sentence level semantics leverage Frame-to-Frame relations to model relationships among sentences. Extensive experiments on two benchmark corpus CNN/DM and NYT demonstrate that our model outperforms six state-of-the-art methods significantly.

Abstractive summarization quality had large improvements since recent language pretraining techniques. However, currently there is a lack of datasets for the growing needs of conversation summarization applications. Thus we collected ForumSum, a diverse and high-quality conversation summarization dataset with human written summaries. The conversations in ForumSum dataset are collected from a wide variety of internet forums. To make the dataset easily expandable, we also release the process of dataset creation. Our experiments show that models trained on ForumSum have better zero-shot and few-shot transferability to other datasets than the existing large chat summarization dataset SAMSum. We also show that using a conversational corpus for pre-training improves the quality of the chat summarization model.