NAACL 2019 Papers

Question-answering plays an important role in e-commerce as it allows potential customers to actively seek crucial information about products or services to help their purchase decision making. Inspired by the recent success of machine reading comprehension (MRC) on formal documents, this paper explores the potential of turning customer reviews into a large source of knowledge that can be exploited to answer user questions. We call this problem Review Reading Comprehension (RRC). To the best of our knowledge, no existing work has been done on RRC. In this work, we first build an RRC dataset called ReviewRC based on a popular benchmark for aspect-based sentiment analysis. Since ReviewRC has limited training examples for RRC (and also for aspect-based sentiment analysis), we then explore a novel post-training approach on the popular language model BERT to enhance the performance of fine-tuning of BERT for RRC. To show the generality of the approach, the proposed post-training is also applied to some other review-based tasks such as aspect extraction and aspect sentiment classification in aspect-based sentiment analysis. Experimental results demonstrate that the proposed post-training is highly effective.

We compare three new datasets for question answering: SQuAD 2.0, QuAC, and CoQA, along several of their new features: (1) unanswerable questions, (2) multi-turn interactions, and (3) abstractive answers. We show that the datasets provide complementary coverage of the first two aspects, but weak coverage of the third. Because of the datasets’ structural similarity, a single extractive model can be easily adapted to any of the datasets and we show improved baseline results on both SQuAD 2.0 and CoQA. Despite the similarity, models trained on one dataset are ineffective on another dataset, but we find moderate performance improvement through pretraining. To encourage cross-evaluation, we release code for conversion between datasets.

Most research in reading comprehension has focused on answering questions based on individual documents or even single paragraphs. We introduce a neural model which integrates and reasons relying on information spread within documents and across multiple documents. We frame it as an inference problem on a graph. Mentions of entities are nodes of this graph while edges encode relations between different mentions (e.g., within- and cross-document co-reference). Graph convolutional networks (GCNs) are applied to these graphs and trained to perform multi-step reasoning. Our Entity-GCN method is scalable and compact, and it achieves state-of-the-art results on a multi-document question answering dataset, WikiHop (Welbl et al., 2018).

Open Information Extraction (OpenIE) extracts meaningful structured tuples from free-form text. Most previous work on OpenIE considers extracting data from one sentence at a time. We describe NeurON, a system for extracting tuples from question-answer pairs. One of the main motivations for NeurON is to be able to extend knowledge bases in a way that considers precisely the information that users care about. NeurON addresses several challenges. First, an answer text is often hard to understand without knowing the question, and second, relevant information can span multiple sentences. To address these, NeurON formulates extraction as a multi-source sequence-to-sequence learning task, wherein it combines distributed representations of a question and an answer to generate knowledge facts. We describe experiments on two real-world datasets that demonstrate that NeurON can find a significant number of new and interesting facts to extend a knowledge base compared to state-of-the-art OpenIE methods.

Generating texts which express complex ideas spanning multiple sentences requires a structured representation of their content (document plan), but these representations are prohibitively expensive to manually produce. In this work, we address the problem of generating coherent multi-sentence texts from the output of an information extraction system, and in particular a knowledge graph. Graphical knowledge representations are ubiquitous in computing, but pose a significant challenge for text generation techniques due to their non-hierarchical nature, collapsing of long-distance dependencies, and structural variety. We introduce a novel graph transforming encoder which can leverage the relational structure of such knowledge graphs without imposing linearization or hierarchical constraints. Incorporated into an encoder-decoder setup, we provide an end-to-end trainable system for graph-to-text generation that we apply to the domain of scientific text. Automatic and human evaluations show that our technique produces more informative texts which exhibit better document structure than competitive encoder-decoder methods.

Recent approaches to question generation have used modifications to a Seq2Seq architecture inspired by advances in machine translation. Models are trained using teacher forcing to optimise only the one-step-ahead prediction. However, at test time, the model is asked to generate a whole sequence, causing errors to propagate through the generation process (exposure bias). A number of authors have suggested that optimising for rewards less tightly coupled to the training data might counter this mismatch. We therefore optimise directly for various objectives beyond simply replicating the ground truth questions, including a novel approach using an adversarial discriminator that seeks to generate questions that are indistinguishable from real examples. We confirm that training with policy gradient methods leads to increases in the metrics used as rewards. We perform a human evaluation, and show that although these metrics have previously been assumed to be good proxies for question quality, they are poorly aligned with human judgement and the model simply learns to exploit the weaknesses of the reward source.

Data-to-text generation can be conceptually divided into two parts: ordering and structuring the information (planning), and generating fluent language describing the information (realization). Modern neural generation systems conflate these two steps into a single end-to-end differentiable system. We propose to split the generation process into a symbolic text-planning stage that is faithful to the input, followed by a neural generation stage that focuses only on realization. For training a plan-to-text generator, we present a method for matching reference texts to their corresponding text plans. For inference time, we describe a method for selecting high-quality text plans for new inputs. We implement and evaluate our approach on the WebNLG benchmark. Our results demonstrate that decoupling text planning from neural realization indeed improves the system’s reliability and adequacy while maintaining fluent output. We observe improvements both in BLEU scores and in manual evaluations. Another benefit of our approach is the ability to output diverse realizations of the same input, paving the way to explicit control over the generated text structure.

We propose neural models to generate high-quality text from structured representations based on Minimal Recursion Semantics (MRS). MRS is a rich semantic representation that encodes more precise semantic detail than other representations such as Abstract Meaning Representation (AMR). We show that a sequence-to-sequence model that maps a linearization of Dependency MRS, a graph-based representation of MRS, to text can achieve a BLEU score of 66.11 when trained on gold data. The performance of the model can be improved further using a high-precision, broad coverage grammar-based parser to generate a large silver training corpus, achieving a final BLEU score of 77.17 on the full test set, and 83.37 on the subset of test data most closely matching the silver data domain. Our results suggest that MRS-based representations are a good choice for applications that need both structured semantics and the ability to produce natural language text as output.

Text generation with generative adversarial networks (GANs) can be divided into the text-based and code-based categories according to the type of signals used for discrimination. In this work, we introduce a novel text-based approach called Soft-GAN to effectively exploit GAN setup for text generation. We demonstrate how autoencoders (AEs) can be used for providing a continuous representation of sentences, which we will refer to as soft-text. This soft representation will be used in GAN discrimination to synthesize similar soft-texts. We also propose hybrid latent code and text-based GAN (LATEXT-GAN) approaches with one or more discriminators, in which a combination of the latent code and the soft-text is used for GAN discriminations. We perform a number of subjective and objective experiments on two well-known datasets (SNLI and Image COCO) to validate our techniques. We discuss the results using several evaluation metrics and show that the proposed techniques outperform the traditional GAN-based text-generation methods.

Generative Adversarial Networks (GANs) are a promising approach for text generation that, unlike traditional language models (LM), does not suffer from the problem of “exposure bias”. However, A major hurdle for understanding the potential of GANs for text generation is the lack of a clear evaluation metric. In this work, we propose to approximate the distribution of text generated by a GAN, which permits evaluating them with traditional probability-based LM metrics. We apply our approximation procedure on several GAN-based models and show that they currently perform substantially worse than state-of-the-art LMs. Our evaluation procedure promotes better understanding of the relation between GANs and LMs, and can accelerate progress in GAN-based text generation.

Learning to hash via generative model has become a powerful paradigm for fast similarity search in documents retrieval. To get binary representation (i.e., hash codes), the discrete distribution prior (i.e., Bernoulli Distribution) is applied to train the variational autoencoder (VAE). However, the discrete stochastic layer is usually incompatible with the backpropagation in the training stage, and thus causes a gradient flow problem because of non-differentiable operators. The reparameterization trick of sampling from a discrete distribution usually inc non-differentiable operators. In this paper, we propose a method, Doc2hash, that solves the gradient flow problem of the discrete stochastic layer by using continuous relaxation on priors, and trains the generative model in an end-to-end manner to generate hash codes. In qualitative and quantitative experiments, we show the proposed model outperforms other state-of-art methods.

Scholars in inter-disciplinary fields like the Digital Humanities are increasingly interested in semantic annotation of specialized corpora. Yet, under-resourced languages, imperfect or noisily structured data, and user-specific classification tasks make it difficult to meet their needs using off-the-shelf models. Manual annotation of large corpora from scratch, meanwhile, can be prohibitively expensive. Thus, we propose an active learning solution for named entity recognition, attempting to maximize a custom model’s improvement per additional unit of manual annotation. Our system robustly handles any domain or user-defined label set and requires no external resources, enabling quality named entity recognition for Humanities corpora where such resources are not available. Evaluating on typologically disparate languages and datasets, we reduce required annotation by 20-60% and greatly outperform a competitive active learning baseline.

The recently released FEVER dataset provided benchmark results on a fact-checking task in which given a factual claim, the system must extract textual evidence (sets of sentences from Wikipedia pages) that support or refute the claim. In this paper, we present a completely task-agnostic pipelined system, AttentiveChecker, consisting of three homogeneous Bi-Directional Attention Flow (BIDAF) networks, which are multi-layer hierarchical networks that represent the context at different levels of granularity. We are the first to apply to this task a bi-directional attention flow mechanism to obtain a query-aware context representation without early summarization. AttentiveChecker can be used to perform document retrieval, sentence selection, and claim verification. Experiments on the FEVER dataset indicate that AttentiveChecker is able to achieve the state-of-the-art results on the FEVER test set.

This paper explores the problem of ranking short social media posts with respect to user queries using neural networks. Instead of starting with a complex architecture, we proceed from the bottom up and examine the effectiveness of a simple, word-level Siamese architecture augmented with attention-based mechanisms for capturing semantic “soft” matches between query and post tokens. Extensive experiments on datasets from the TREC Microblog Tracks show that our simple models not only achieve better effectiveness than existing approaches that are far more complex or exploit a more diverse set of relevance signals, but are also much faster.

In Community-based Question Answering system(CQA), Answer Selection(AS) is a critical task, which focuses on finding a suitable answer within a list of candidate answers. For neural network models, the key issue is how to model the representations of QA text pairs and calculate the interactions between them. We propose a Sequential Attention with Keyword Mask model(SAKM) for CQA to imitate human reading behavior. Question and answer text regard each other as context within keyword-mask attention when encoding the representations, and repeat multiple times(hops) in a sequential style. So the QA pairs capture features and information from both question text and answer text, interacting and improving vector representations iteratively through hops. The flexibility of the model allows to extract meaningful keywords from the sentences and enhance diverse mutual information. We perform on answer selection tasks and multi-level answer ranking tasks. Experiment results demonstrate the superiority of our proposed model on community-based QA datasets.

For many structured learning tasks, the data annotation process is complex and costly. Existing annotation schemes usually aim at acquiring completely annotated structures, under the common perception that partial structures are of low quality and could hurt the learning process. This paper questions this common perception, motivated by the fact that structures consist of interdependent sets of variables. Thus, given a fixed budget, partly annotating each structure may provide the same level of supervision, while allowing for more structures to be annotated. We provide an information theoretic formulation for this perspective and use it, in the context of three diverse structured learning tasks, to show that learning from partial structures can sometimes outperform learning from complete ones. Our findings may provide important insights into structured data annotation schemes and could support progress in learning protocols for structured tasks.

In this paper, we introduce an embedding model, named CapsE, exploring a capsule network to model relationship triples (subject, relation, object). Our CapsE represents each triple as a 3-column matrix where each column vector represents the embedding of an element in the triple. This 3-column matrix is then fed to a convolution layer where multiple filters are operated to generate different feature maps. These feature maps are reconstructed into corresponding capsules which are then routed to another capsule to produce a continuous vector. The length of this vector is used to measure the plausibility score of the triple. Our proposed CapsE obtains better performance than previous state-of-the-art embedding models for knowledge graph completion on two benchmark datasets WN18RR and FB15k-237, and outperforms strong search personalization baselines on SEARCH17.

Several datasets have recently been constructed to expose brittleness in models trained on existing benchmarks. While model performance on these challenge datasets is significantly lower compared to the original benchmark, it is unclear what particular weaknesses they reveal. For example, a challenge dataset may be difficult because it targets phenomena that current models cannot capture, or because it simply exploits blind spots in a model’s specific training set. We introduce inoculation by fine-tuning, a new analysis method for studying challenge datasets by exposing models (the metaphorical patient) to a small amount of data from the challenge dataset (a metaphorical pathogen) and assessing how well they can adapt. We apply our method to analyze the NLI “stress tests” (Naik et al., 2018) and the Adversarial SQuAD dataset (Jia and Liang, 2017). We show that after slight exposure, some of these datasets are no longer challenging, while others remain difficult. Our results indicate that failures on challenge datasets may lead to very different conclusions about models, training datasets, and the challenge datasets themselves.

We present a resource for the task of FrameNet semantic frame disambiguation of over 5,000 word-sentence pairs from the Wikipedia corpus. The annotations were collected using a novel crowdsourcing approach with multiple workers per sentence to capture inter-annotator disagreement. In contrast to the typical approach of attributing the best single frame to each word, we provide a list of frames with disagreement-based scores that express the confidence with which each frame applies to the word. This is based on the idea that inter-annotator disagreement is at least partly caused by ambiguity that is inherent to the text and frames. We have found many examples where the semantics of individual frames overlap sufficiently to make them acceptable alternatives for interpreting a sentence. We have argued that ignoring this ambiguity creates an overly arbitrary target for training and evaluating natural language processing systems - if humans cannot agree, why would we expect the correct answer from a machine to be any different? To process this data we also utilized an expanded lemma-set provided by the Framester system, which merges FN with WordNet to enhance coverage. Our dataset includes annotations of 1,000 sentence-word pairs whose lemmas are not part of FN. Finally we present metrics for evaluating frame disambiguation systems that account for ambiguity.