EMNLP 2019 Papers

Existing recurrent neural language models often fail to capture higher-level structure present in text: for example, rhyming patterns present in poetry. Much prior work on poetry generation uses manually defined constraints which are satisfied during decoding using either specialized decoding procedures or rejection sampling. The rhyming constraints themselves are typically not learned by the generator. We propose an alternate approach that uses a structured discriminator to learn a poetry generator that directly captures rhyming constraints in a generative adversarial setup. By causing the discriminator to compare poems based only on a learned similarity matrix of pairs of line ending words, the proposed approach is able to successfully learn rhyming patterns in two different English poetry datasets (Sonnet and Limerick) without explicitly being provided with any phonetic information

Sentence position is a strong feature for news summarization, since the lead often (but not always) summarizes the key points of the article. In this paper, we show that recent neural systems excessively exploit this trend, which although powerful for many inputs, is also detrimental when summarizing documents where important content should be extracted from later parts of the article. We propose two techniques to make systems sensitive to the importance of content in different parts of the article. The first technique employs ‘unbiased’ data; i.e., randomly shuffled sentences of the source document, to pretrain the model. The second technique uses an auxiliary ROUGE-based loss that encourages the model to distribute importance scores throughout a document by mimicking sentence-level ROUGE scores on the training data. We show that these techniques significantly improve the performance of a competitive reinforcement learning based extractive system, with the auxiliary loss being more powerful than pretraining.

Generating paraphrases from given sentences involves decoding words step by step from a large vocabulary. To learn a decoder, supervised learning which maximizes the likelihood of tokens always suffers from the exposure bias. Although both reinforcement learning (RL) and imitation learning (IL) have been widely used to alleviate the bias, the lack of direct comparison leads to only a partial image on their benefits. In this work, we present an empirical study on how RL and IL can help boost the performance of generating paraphrases, with the pointer-generator as a base model. Experiments on the benchmark datasets show that (1) imitation learning is constantly better than reinforcement learning; and (2) the pointer-generator models with imitation learning outperform the state-of-the-art methods with a large margin.

We propose SUM-QE, a novel Quality Estimation model for summarization based on BERT. The model addresses linguistic quality aspects that are only indirectly captured by content-based approaches to summary evaluation, without involving comparison with human references. SUM-QE achieves very high correlations with human ratings, outperforming simpler models addressing these linguistic aspects. Predictions of the SUM-QE model can be used for system development, and to inform users of the quality of automatically produced summaries and other types of generated text.

We present a novel approach for generating poetry automatically for the morphologically rich Finnish language by using a genetic algorithm. The approach improves the state of the art of the previous Finnish poem generators by introducing a higher degree of freedom in terms of structural creativity. Our approach is evaluated and described within the paradigm of computational creativity, where the fitness functions of the genetic algorithm are assimilated with the notion of aesthetics. The output is considered to be a poem 81.5% of the time by human evaluators.

Text summarization is considered as a challenging task in the NLP community. The availability of datasets for the task of multilingual text summarization is rare, and such datasets are difficult to construct. In this work, we build an abstract text summarizer for the German language text using the state-of-the-art “Transformer” model. We propose an iterative data augmentation approach which uses synthetic data along with the real summarization data for the German language. To generate synthetic data, the Common Crawl (German) dataset is exploited, which covers different domains. The synthetic data is effective for the low resource condition and is particularly helpful for our multilingual scenario where availability of summarizing data is still a challenging issue. The data are also useful in deep learning scenarios where the neural models require a large amount of training data for utilization of its capacity. The obtained summarization performance is measured in terms of ROUGE and BLEU score. We achieve an absolute improvement of +1.5 and +16.0 in ROUGE1 F1 (R1_F1) on the development and test sets, respectively, compared to the system which does not rely on data augmentation.

Stories generated with neural language models have shown promise in grammatical and stylistic consistency. However, the generated stories are still lacking in common sense reasoning, e.g., they often contain sentences deprived of world knowledge. We propose a simple multi-task learning scheme to achieve quantitatively better common sense reasoning in language models by leveraging auxiliary training signals from datasets designed to provide common sense grounding. When combined with our two-stage fine-tuning pipeline, our method achieves improved common sense reasoning and state-of-the-art perplexity on the WritingPrompts (Fan et al., 2018) story generation dataset.

The neural seq2seq based question generation (QG) is prone to generating generic and undiversified questions that are poorly relevant to the given passage and target answer. In this paper, we propose two methods to address the issue. (1) By a partial copy mechanism, we prioritize words that are morphologically close to words in the input passage when generating questions; (2) By a QA-based reranker, from the n-best list of question candidates, we select questions that are preferred by both the QA and QG model. Experiments and analyses demonstrate that the proposed two methods substantially improve the relevance of generated questions to passages and answers.

Existing approaches to recipe generation are unable to create recipes for users with culinary preferences but incomplete knowledge of ingredients in specific dishes. We propose a new task of personalized recipe generation to help these users: expanding a name and incomplete ingredient details into complete natural-text instructions aligned with the user’s historical preferences. We attend on technique- and recipe-level representations of a user’s previously consumed recipes, fusing these ‘user-aware’ representations in an attention fusion layer to control recipe text generation. Experiments on a new dataset of 180K recipes and 700K interactions show our model’s ability to generate plausible and personalized recipes compared to non-personalized baselines.

Search applications often display shortened sentences which must contain certain query terms and must fit within the space constraints of a user interface. This work introduces a new transition-based sentence compression technique developed for such settings. Our query-focused method constructs length and lexically constrained compressions in linear time, by growing a subgraph in the dependency parse of a sentence. This theoretically efficient approach achieves an 11x empirical speedup over baseline ILP methods, while better reconstructing gold constrained shortenings. Such speedups help query-focused applications, because users are measurably hindered by interface lags. Additionally, our technique does not require an ILP solver or a GPU.

Paraphrase generation is an interesting and challenging NLP task which has numerous practical applications. In this paper, we analyze datasets commonly used for paraphrase generation research, and show that simply parroting input sentences surpasses state-of-the-art models in the literature when evaluated on standard metrics. Our findings illustrate that a model could be seemingly adept at generating paraphrases, despite only making trivial changes to the input sentence or even none at all.

Answer selection is an important research problem, with applications in many areas. Previous deep learning based approaches for the task mainly adopt the Compare-Aggregate architecture that performs word-level comparison followed by aggregation. In this work, we take a departure from the popular Compare-Aggregate architecture, and instead, propose a new gated self-attention memory network for the task. Combined with a simple transfer learning technique from a large-scale online corpus, our model outperforms previous methods by a large margin, achieving new state-of-the-art results on two standard answer selection datasets: TrecQA and WikiQA.

Reading comprehension models have been successfully applied to extractive text answers, but it is unclear how best to generalize these models to abstractive numerical answers. We enable a BERT-based reading comprehension model to perform lightweight numerical reasoning. We augment the model with a predefined set of executable ‘programs’ which encompass simple arithmetic as well as extraction. Rather than having to learn to manipulate numbers directly, the model can pick a program and execute it. On the recent Discrete Reasoning Over Passages (DROP) dataset, designed to challenge reading comprehension models, we show a 33% absolute improvement by adding shallow programs. The model can learn to predict new operations when appropriate in a math word problem setting (Roy and Roth, 2015) with very few training examples.

We introduce the first open-domain dataset, called QuaRTz, for reasoning about textual qualitative relationships. QuaRTz contains general qualitative statements, e.g., “A sunscreen with a higher SPF protects the skin longer.”, twinned with 3864 crowdsourced situated questions, e.g., “Billy is wearing sunscreen with a lower SPF than Lucy. Who will be best protected from the sun?”, plus annotations of the properties being compared. Unlike previous datasets, the general knowledge is textual and not tied to a fixed set of relationships, and tests a system’s ability to comprehend and apply textual qualitative knowledge in a novel setting. We find state-of-the-art results are substantially (20%) below human performance, presenting an open challenge to the NLP community.

Because it is not feasible to collect training data for every language, there is a growing interest in cross-lingual transfer learning. In this paper, we systematically explore zero-shot cross-lingual transfer learning on reading comprehension tasks with language representation model pre-trained on multi-lingual corpus. The experimental results show that with pre-trained language representation zero-shot learning is feasible, and translating the source data into the target language is not necessary and even degrades the performance. We further explore what does the model learn in zero-shot setting.

Machine comprehension of texts longer than a single sentence often requires coreference resolution. However, most current reading comprehension benchmarks do not contain complex coreferential phenomena and hence fail to evaluate the ability of models to resolve coreference. We present a new crowdsourced dataset containing more than 24K span-selection questions that require resolving coreference among entities in over 4.7K English paragraphs from Wikipedia. Obtaining questions focused on such phenomena is challenging, because it is hard to avoid lexical cues that shortcut complex reasoning. We deal with this issue by using a strong baseline model as an adversary in the crowdsourcing loop, which helps crowdworkers avoid writing questions with exploitable surface cues. We show that state-of-the-art reading comprehension models perform significantly worse than humans on this benchmark—the best model performance is 70.5 F1, while the estimated human performance is 93.4 F1.

Question answering is an AI-complete problem, but existing datasets lack key elements of language understanding such as coreference and ellipsis resolution. We consider sequential question answering: multiple questions are asked one-by-one in a conversation between a questioner and an answerer. Answering these questions is only possible through understanding the conversation history. We introduce the task of question-in-context rewriting: given the context of a conversation’s history, rewrite a context-dependent into a self-contained question with the same answer. We construct, CANARD, a dataset of 40,527 questions based on QuAC (Choi et al., 2018) and train Seq2Seq models for incorporating context into standalone questions.

In this paper, we establish the effectiveness of using hard negatives, coupled with a siamese network and a suitable loss function, for the tasks of answer selection and answer triggering. We show that the choice of sampling strategy is key for achieving improved performance on these tasks. Evaluating on recent answer selection datasets - InsuranceQA, SelQA, and an internal QA dataset, we show that using hard negatives with relatively simple model architectures (bag of words and LSTM-CNN) drives significant performance gains. On InsuranceQA, this strategy alone improves over previously reported results by a minimum of 1.6 points in P@1. Using hard negatives with a Transformer encoder provides a further improvement of 2.3 points. Further, we propose to use quadruplet loss for answer triggering, with the aim of producing globally meaningful similarity scores. We show that quadruplet loss function coupled with the selection of hard negatives enables bag-of-words models to improve F1 score by 2.3 points over previous baselines, on SelQA answer triggering dataset. Our results provide key insights into answer selection and answer triggering tasks.

We present a novel approach to answering sequential questions based on structured objects such as knowledge bases or tables without using a logical form as an intermediate representation. We encode tables as graphs using a graph neural network model based on the Transformer architecture. The answers are then selected from the encoded graph using a pointer network. This model is appropriate for processing conversations around structured data, where the attention mechanism that selects the answers to a question can also be used to resolve conversational references. We demonstrate the validity of this approach with competitive results on the Sequential Question Answering (SQA) task.