EMNLP 2016 Papers

Language documentation begins by gathering speech. Manual or automatic transcription at the word level is typically not possible because of the absence of an orthography or prior lexicon, and though manual phone-mic transcription is possible, it is prohibitively slow. On the other hand, translations of the minority language into a major language are more easily acquired. We propose a method to harness such translations to improve automatic phoneme recognition. The method assumes no prior lexicon or translation model, instead learning them from phoneme lattices and translations of the speech being transcribed. Experiments demonstrate phoneme error rate improvements against two baselines and the model’s ability to learn useful bilingual lexical entries.

We introduce the first global recursive neural parsing model with optimality guarantees during decoding. To support global features, we give up dynamic programs and instead search directly in the space of all possible subtrees. Although this space is exponentially large in the sentence length, we show it is possible to learn an efficient A* parser. We augment existing parsing models, which have informative bounds on the outside score, with a global model that has loose bounds but only needs to model non-local phenomena. The global model is trained with a new objective that encourages the parser to explore a tiny fraction of the search space. The approach is applied to CCG parsing, improving state-of-the-art accuracy by 0.4 F1. The parser finds the optimal parse for 99.9% of held-out sentences, exploring on average only 190 subtrees.

Most successful information extraction systems operate with access to a large collection of documents. In this work, we explore the task of acquiring and incorporating external evidence to improve extraction accuracy in domains where the amount of training data is scarce. This process entails issuing search queries, extraction from new sources and reconciliation of extracted values, which are repeated until sufficient evidence is collected. We approach the problem using a reinforcement learning framework where our model learns to select optimal actions based on contextual information. We employ a deep Q-network, trained to optimize a reward function that reflects extraction accuracy while penalizing extra effort. Our experiments on two databases -- of shooting incidents, and food adulteration cases -- demonstrate that our system significantly outperforms traditional extractors and a competitive meta-classifier baseline.

Most existing knowledge base (KB) embedding methods solely learn from time-unknown fact triples but neglect the temporal information in the knowledge base. In this paper, we propose a novel time-aware KB embedding approach taking advantage of the happening time of facts. Specifically, we use temporal order constraints to model transformation between time-sensitive relations and enforce the embeddings to be temporally consistent and more accurate. We empirically evaluate our approach in two tasks of link prediction and triple classification. Experimental re-sults show that our method outperforms other baselines on the two tasks consistently.

,

This paper demonstrates that it is possible for a parser to improve its performance with a human in the loop, by posing simple questions to non-experts. For example, given the first sentence of this abstract, if the parser is uncertain about the subject of the verb “pose,” it could generate the question What would pose something? with candidate answers this paper and a parser. Any fluent speaker can answer this question, and the correct answer resolves the original uncertainty. We apply the approach to a CCG parser, converting uncertain attachment decisions into natural language questions about the arguments of verbs. Experiments show that crowd workers can answer these questions quickly, accurately and cheaply. Our human-in-the-loop parser improves on the state of the art with less than 2 questions per sentence on average, with a gain of 1.7 F1 on the 10% of sentences whose parses are changed.

We recast syntactic parsing as a language modeling problem and use recent advances in neural network language modeling to achieve a new state of the art for constituency Penn Treebank parsing — 93.8 F 1 on section 23, us-ing 2-21 as training, 24 as development, plus tri-training. When trees are converted to Stanford dependencies, UAS and LAS are 95.9% and 94.1%.

Morphological segmentation has traditionally been modeled with non-hierarchical models, which yield flat segmentations as output. In many cases, however, proper morphological analysis requires hierarchical structure -- especially in the case of derivational morphology. In this work, we introduce a discriminative, joint model of morphological segmentation along with the orthographic changes that occur during word formation. To the best of our knowledge, this is the first attempt to approach discriminative segmentation with a context-free model. Additionally, we release an annotated treebank of 7454 English words with constituency parses, encouraging future research in this area.

Recent work on word ordering has argued that syntactic structure is important, or even required, for effectively recovering the order of a sentence. We find that, in fact, an n-gram language model with a simple heuristic gives strong results on this task. Furthermore, we show that a long short-term memory (LSTM) language model is even more effective at recovering order, with our basic model outperforming a state-of-the-art syntactic model by 11.5 BLEU points. Additional data and larger beams yield further gains, at the expense of training and search time.

We introduce a novel approach to the decoding problem in transition-based parsing: heuristic backtracking. This algorithm uses a series of partial parses on the sentence to locate the best candidate parse, using confidence estimates of transition decisions as a heuristic to guide the starting points of the search. This allows us to achieve a parse accuracy comparable to beam search, despite using fewer transitions. When used to augment a Stack-LSTM transition-based parser, the parser shows an unlabeled attachment score of up to 93.30% for English and 87.61% for Chinese.

To alleviate the shortage of labeled data, we propose to use bilingually-constrained synthetic implicit data for implicit discourse relation recognition. These data are extracted from a bilingual sentence-aligned corpus according to the implicit/explicit mismatch be-tween different languages. Incorporating these data via a multi-task neural network model achieves significant improvements over baselines, on both the English PDTB and Chinese CDTB data sets.

Open information extraction (Open IE) was presented as an unrestricted variant of traditional information extraction. It has been gaining substantial attention, manifested by a large number of automatic Open IE extractors and downstream applications. In spite of this broad attention, the Open IE task definition has been lacking – there are no formal guidelines and no large scale gold standard annotation. Subsequently, the various implementations of Open IE resorted to small scale post-hoc evaluations, inhibiting an objective and re-producible cross-system comparison. In this work, we develop a methodology that leverages the recent QA-SRL annotation to create a first independent and large scale Open IE annotation, 1 and use it to automatically compare the most prominent Open IE systems.

In this paper we discuss a process for quantifying the behavioral impact of a domain-customized machine translation system deployed on a large-scale e-commerce platform. We discuss several machine translation systems that we trained using aligned text from product listing descriptions written in multiple languages. We document the quality improvements of these systems as measured through automated quality measures and crowdsourced human quality assessments. We then measure the effect of these quality improvements on user behavior using an automated A/B testing framework. Through testing we observed an increase in key e-commerce metrics, including a significant increase in purchases.

Mapping word embeddings of different languages into a single space has multiple applications. In order to map from a source space into a target space, a common approach is to learn a linear mapping that minimizes the distances between equivalences listed in a bilingual dictionary. In this paper, we propose a framework that generalizes previous work, provides an efficient exact method to learn the optimal linear transformation and yields the best bilingual results in translation induction while preserving monolingual performance in an analogy task.

In this paper, we improve the attention or alignment accuracy of neural machine translation by utilizing the alignments of training sentence pairs. We simply compute the distance between the machine attentions and the "true" alignments, and minimize this cost in the training procedure. Our experiments on large-scale Chinese-to-English task show that our model improves both translation and alignment qualities significantly over the large-vocabulary neural machine translation system, and even beats a state-of-the-art traditional syntax-based system.

We investigate how neural, encoder-decoder translation systems output target strings of appropriate lengths, finding that a collection of hidden units learns to explicitly implement this functionality.

Dependency tree-to-tree translation models are powerful because they can naturally handle long range reorderings which is important for distant language pairs. The translation process is easy if it can be accomplished only by replacing non-terminals in translation rules with other rules. However it is sometimes necessary to adjoin translation rules. Flex-ible non-terminals have been proposed as a promising solution for this problem. A flex-ible non-terminal provides several insertion position candidates for the rules to be ad-joined, but it increases the computational cost of decoding. In this paper we propose a neural network based insertion position selection model to reduce the computational cost by selecting the appropriate insertion positions. The experimental results show the proposed model can select the appropriate insertion position with a high accuracy. It reduces the decoding time and improves the translation quality owing to reduced search space.

Discourse parsing is considered as one of the most challenging natural language processing (NLP) tasks. Implicit discourse relation clas-sification is the bottleneck for discourse parsing. Without the guide of explicit discourse connectives, the relation of sentence pairs are very hard to be inferred. This paper proposes a stacking neural network model to solve the classification problem in which a convolutional neural network (CNN) is utilized for sentence modeling and a collaborative gated neural network (CGNN) is proposed for feature transformation. Our evaluation and comparisons show that the proposed model outperforms previous state-of-the-art systems.

Coreference resolution systems are typically trained with heuristic loss functions that require careful tuning. In this paper we instead apply reinforcement learning to directly optimize a neural mention-ranking model for coreference evaluation metrics. We experiment with two approaches: the REINFORCE policy gradient algorithm and a reward-rescaled max-margin objective. We find the latter to be more effective, resulting in significant improvements over the current state-of-the-art on the English and Chinese portions of the CoNLL 2012 Shared Task.