EMNLP 2018 Papers

Causal understanding is essential for many kinds of decision-making, but causal inference from observational data has typically only been applied to structured, low-dimensional datasets. While text classifiers produce low-dimensional outputs, their use in causal inference has not previously been studied. To facilitate causal analyses based on language data, we consider the role that text classifiers can play in causal inference through established modeling mechanisms from the causality literature on missing data and measurement error. We demonstrate how to conduct causal analyses using text classifiers on simulated and Yelp data, and discuss the opportunities and challenges of future work that uses text data in causal inference.

Prevalence estimation is the task of inferring the relative frequency of classes of unlabeled examples in a group—for example, the proportion of a document collection with positive sentiment. Previous work has focused on aggregating and adjusting discriminative individual classifiers to obtain prevalence point estimates. But imperfect classifier accuracy ought to be reflected in uncertainty over the predicted prevalence for scientifically valid inference. In this work, we present (1) a generative probabilistic modeling approach to prevalence estimation, and (2) the construction and evaluation of prevalence confidence intervals; in particular, we demonstrate that an off-the-shelf discriminative classifier can be given a generative re-interpretation, by backing out an implicit individual-level likelihood function, which can be used to conduct fast and simple group-level Bayesian inference. Empirically, we demonstrate our approach provides better confidence interval coverage than an alternative, and is dramatically more robust to shifts in the class prior between training and testing.

Multi-task learning has an ability to share the knowledge among related tasks and implicitly increase the training data. However, it has long been frustrated by the interference among tasks. This paper investigates the performance of capsule network for text, and proposes a capsule-based multi-task learning architecture, which is unified, simple and effective. With the advantages of capsules for feature clustering, proposed task routing algorithm can cluster the features for each task in the network, which helps reduce the interference among tasks. Experiments on six text classification datasets demonstrate the effectiveness of our models and their characteristics for feature clustering.

We propose a novel model for multi-label text classification, which is based on sequence-to-sequence learning. The model generates higher-level semantic unit representations with multi-level dilated convolution as well as a corresponding hybrid attention mechanism that extracts both the information at the word-level and the level of the semantic unit. Our designed dilated convolution effectively reduces dimension and supports an exponential expansion of receptive fields without loss of local information, and the attention-over-attention mechanism is able to capture more summary relevant information from the source context. Results of our experiments show that the proposed model has significant advantages over the baseline models on the dataset RCV1-V2 and Ren-CECps, and our analysis demonstrates that our model is competitive to the deterministic hierarchical models and it is more robust to classifying low-frequency labels

Multi-task learning in text classification leverages implicit correlations among related tasks to extract common features and yield performance gains. However, a large body of previous work treats labels of each task as independent and meaningless one-hot vectors, which cause a loss of potential label information. In this paper, we propose Multi-Task Label Embedding to convert labels in text classification into semantic vectors, thereby turning the original tasks into vector matching tasks. Our model utilizes semantic correlations among tasks and makes it convenient to scale or transfer when new tasks are involved. Extensive experiments on five benchmark datasets for text classification show that our model can effectively improve the performances of related tasks with semantic representations of labels and additional information from each other.

Clustering news across languages enables efficient media monitoring by aggregating articles from multilingual sources into coherent stories. Doing so in an online setting allows scalable processing of massive news streams. To this end, we describe a novel method for clustering an incoming stream of multilingual documents into monolingual and crosslingual clusters. Unlike typical clustering approaches that report results on datasets with a small and known number of labels, we tackle the problem of discovering an ever growing number of cluster labels in an online fashion, using real news datasets in multiple languages. In our formulation, the monolingual clusters group together documents while the crosslingual clusters group together monolingual clusters, one per language that appears in the stream. Our method is simple to implement, computationally efficient and produces state-of-the-art results on datasets in German, English and Spanish.

While the celebrated Word2Vec technique yields semantically rich representations for individual words, there has been relatively less success in extending to generate unsupervised sentences or documents embeddings. Recent work has demonstrated that a distance measure between documents called Word Mover’s Distance (WMD) that aligns semantically similar words, yields unprecedented KNN classification accuracy. However, WMD is expensive to compute, and it is hard to extend its use beyond a KNN classifier. In this paper, we propose the Word Mover’s Embedding (WME), a novel approach to building an unsupervised document (sentence) embedding from pre-trained word embeddings. In our experiments on 9 benchmark text classification datasets and 22 textual similarity tasks, the proposed technique consistently matches or outperforms state-of-the-art techniques, with significantly higher accuracy on problems of short length.

In order to learn universal sentence representations, previous methods focus on complex recurrent neural networks or supervised learning. In this paper, we propose a mean-max attention autoencoder (mean-max AAE) within the encoder-decoder framework. Our autoencoder rely entirely on the MultiHead self-attention mechanism to reconstruct the input sequence. In the encoding we propose a mean-max strategy that applies both mean and max pooling operations over the hidden vectors to capture diverse information of the input. To enable the information to steer the reconstruction process dynamically, the decoder performs attention over the mean-max representation. By training our model on a large collection of unlabelled data, we obtain high-quality representations of sentences. Experimental results on a broad range of 10 transfer tasks demonstrate that our model outperforms the state-of-the-art unsupervised single methods, including the classical skip-thoughts and the advanced skip-thoughts+LN model. Furthermore, compared with the traditional recurrent neural network, our mean-max AAE greatly reduce the training time.

A hallmark of variational autoencoders (VAEs) for text processing is their combination of powerful encoder-decoder models, such as LSTMs, with simple latent distributions, typically multivariate Gaussians. These models pose a difficult optimization problem: there is an especially bad local optimum where the variational posterior always equals the prior and the model does not use the latent variable at all, a kind of “collapse” which is encouraged by the KL divergence term of the objective. In this work, we experiment with another choice of latent distribution, namely the von Mises-Fisher (vMF) distribution, which places mass on the surface of the unit hypersphere. With this choice of prior and posterior, the KL divergence term now only depends on the variance of the vMF distribution, giving us the ability to treat it as a fixed hyperparameter. We show that doing so not only averts the KL collapse, but consistently gives better likelihoods than Gaussians across a range of modeling conditions, including recurrent language modeling and bag-of-words document modeling. An analysis of the properties of our vMF representations shows that they learn richer and more nuanced structures in their latent representations than their Gaussian counterparts.

Learning a matching function between two text sequences is a long standing problem in NLP research. This task enables many potential applications such as question answering and paraphrase identification. This paper proposes Co-Stack Residual Affinity Networks (CSRAN), a new and universal neural architecture for this problem. CSRAN is a deep architecture, involving stacked (multi-layered) recurrent encoders. Stacked/Deep architectures are traditionally difficult to train, due to the inherent weaknesses such as difficulty with feature propagation and vanishing gradients. CSRAN incorporates two novel components to take advantage of the stacked architecture. Firstly, it introduces a new bidirectional alignment mechanism that learns affinity weights by fusing sequence pairs across stacked hierarchies. Secondly, it leverages a multi-level attention refinement component between stacked recurrent layers. The key intuition is that, by leveraging information across all network hierarchies, we can not only improve gradient flow but also improve overall performance. We conduct extensive experiments on six well-studied text sequence matching datasets, achieving state-of-the-art performance on all.

Pseudo relevance feedback (PRF) is commonly used to boost the performance of traditional information retrieval (IR) models by using top-ranked documents to identify and weight new query terms, thereby reducing the effect of query-document vocabulary mismatches. While neural retrieval models have recently demonstrated strong results for ad-hoc retrieval, combining them with PRF is not straightforward due to incompatibilities between existing PRF approaches and neural architectures. To bridge this gap, we propose an end-to-end neural PRF framework that can be used with existing neural IR models by embedding different neural models as building blocks. Extensive experiments on two standard test collections confirm the effectiveness of the proposed NPRF framework in improving the performance of two state-of-the-art neural IR models.

Common recurrent neural architectures scale poorly due to the intrinsic difficulty in parallelizing their state computations. In this work, we propose the Simple Recurrent Unit (SRU), a light recurrent unit that balances model capacity and scalability. SRU is designed to provide expressive recurrence, enable highly parallelized implementation, and comes with careful initialization to facilitate training of deep models. We demonstrate the effectiveness of SRU on multiple NLP tasks. SRU achieves 5—9x speed-up over cuDNN-optimized LSTM on classification and question answering datasets, and delivers stronger results than LSTM and convolutional models. We also obtain an average of 0.7 BLEU improvement over the Transformer model (Vaswani et al., 2017) on translation by incorporating SRU into the architecture.

We introduce a novel type of text representation that preserves the 2D layout of a document. This is achieved by encoding each document page as a two-dimensional grid of characters. Based on this representation, we present a generic document understanding pipeline for structured documents. This pipeline makes use of a fully convolutional encoder-decoder network that predicts a segmentation mask and bounding boxes. We demonstrate its capabilities on an information extraction task from invoices and show that it significantly outperforms approaches based on sequential text or document images.

Self-attention networks have proven to be of profound value for its strength of capturing global dependencies. In this work, we propose to model localness for self-attention networks, which enhances the ability of capturing useful local context. We cast localness modeling as a learnable Gaussian bias, which indicates the central and scope of the local region to be paid more attention. The bias is then incorporated into the original attention distribution to form a revised distribution. To maintain the strength of capturing long distance dependencies while enhance the ability of capturing short-range dependencies, we only apply localness modeling to lower layers of self-attention networks. Quantitative and qualitative analyses on Chinese-English and English-German translation tasks demonstrate the effectiveness and universality of the proposed approach.

We present LSTM-Shuttle, which applies human speed reading techniques to natural language processing tasks for accurate and efficient comprehension. In contrast to previous work, LSTM-Shuttle not only reads shuttling forward but also goes back. Shuttling forward enables high efficiency, and going backward gives the model a chance to recover lost information, ensuring better prediction. We evaluate LSTM-Shuttle on sentiment analysis, news classification, and cloze on IMDB, Rotten Tomatoes, AG, and Children’s Book Test datasets. We show that LSTM-Shuttle predicts both better and more quickly. To demonstrate how LSTM-Shuttle actually behaves, we also analyze the shuttling operation and present a case study.

Character-level string-to-string transduction is an important component of various NLP tasks. The goal is to map an input string to an output string, where the strings may be of different lengths and have characters taken from different alphabets. Recent approaches have used sequence-to-sequence models with an attention mechanism to learn which parts of the input string the model should focus on during the generation of the output string. Both soft attention and hard monotonic attention have been used, but hard non-monotonic attention has only been used in other sequence modeling tasks and has required a stochastic approximation to compute the gradient. In this work, we introduce an exact, polynomial-time algorithm for marginalizing over the exponential number of non-monotonic alignments between two strings, showing that hard attention models can be viewed as neural reparameterizations of the classical IBM Model 1. We compare soft and hard non-monotonic attention experimentally and find that the exact algorithm significantly improves performance over the stochastic approximation and outperforms soft attention.

Activation functions play a crucial role in neural networks because they are the nonlinearities which have been attributed to the success story of deep learning. One of the currently most popular activation functions is ReLU, but several competitors have recently been proposed or ‘discovered’, including LReLU functions and swish. While most works compare newly proposed activation functions on few tasks (usually from image classification) and against few competitors (usually ReLU), we perform the first largescale comparison of 21 activation functions across eight different NLP tasks. We find that a largely unknown activation function performs most stably across all tasks, the so-called penalized tanh function. We also show that it can successfully replace the sigmoid and tanh gates in LSTM cells, leading to a 2 percentage point (pp) improvement over the standard choices on a challenging NLP task.

Vulgar words are employed in language use for several different functions, ranging from expressing aggression to signaling group identity or the informality of the communication. This versatility of usage of a restricted set of words is challenging for downstream applications and has yet to be studied quantitatively or using natural language processing techniques. We introduce a novel data set of 7,800 tweets from users with known demographic traits where all instances of vulgar words are annotated with one of the six categories of vulgar word use. Using this data set, we present the first analysis of the pragmatic aspects of vulgarity and how they relate to social factors. We build a model able to predict the category of a vulgar word based on the immediate context it appears in with 67.4 macro F1 across six classes. Finally, we demonstrate the utility of modeling the type of vulgar word use in context by using this information to achieve state-of-the-art performance in hate speech detection on a benchmark data set.

This paper presents a set of dimensions to characterize the association between two people. We distinguish between interactions (when somebody refers to somebody in a conversation) and relationships (a sequence of interactions). We work with dialogue scripts from the TV show Friends, and do not impose any restrictions on the interactions and relationships. We introduce and analyze a new corpus, and present experimental results showing that the task can be automated.