EMNLP 2017 Papers

Training a POS tagging model with crosslingual transfer learning usually requires linguistic knowledge and resources about the relation between the source language and the target language. In this paper, we introduce a cross-lingual transfer learning model for POS tagging without ancillary resources such as parallel corpora. The proposed cross-lingual model utilizes a common BLSTM that enables knowledge transfer from other languages, and private BLSTMs for language-specific representations. The cross-lingual model is trained with language-adversarial training and bidirectional language modeling as auxiliary objectives to better represent language-general information while not losing the information about a specific target language. Evaluating on POS datasets from 14 languages in the Universal Dependencies corpus, we show that the proposed transfer learning model improves the POS tagging performance of the target languages without exploiting any linguistic knowledge between the source language and the target language.

In translation, considering the document as a whole can help to resolve ambiguities and inconsistencies. In this paper, we propose a cross-sentence context-aware approach and investigate the influence of historical contextual information on the performance of neural machine translation (NMT). First, this history is summarized in a hierarchical way. We then integrate the historical representation into NMT in two strategies: 1) a warm-start of encoder and decoder states, and 2) an auxiliary context source for updating decoder states. Experimental results on a large Chinese-English translation task show that our approach significantly improves upon a strong attention-based NMT system by up to +2.1 BLEU points.

Attentional sequence-to-sequence models have become the new standard for machine translation, but one challenge of such models is a significant increase in training and decoding cost compared to phrase-based systems. In this work we focus on efficient decoding, with a goal of achieving accuracy close the state-of-the-art in neural machine translation (NMT), while achieving CPU decoding speed/throughput close to that of a phrasal decoder. We approach this problem from two angles: First, we describe several techniques for speeding up an NMT beam search decoder, which obtain a 4.4x speedup over a very efficient baseline decoder without changing the decoder output. Second, we propose a simple but powerful network architecture which uses an RNN (GRU/LSTM) layer at bottom, followed by a series of stacked fully-connected layers applied at every timestep. This architecture achieves similar accuracy to a deep recurrent model, at a small fraction of the training and decoding cost. By combining these techniques, our best system achieves a very competitive accuracy of 38.3 BLEU on WMT English-French NewsTest2014, while decoding at 100 words/sec on single-threaded CPU. We believe this is the best published accuracy/speed trade-off of an NMT system.

Stylistic variations of language, such as formality, carry speakers’ intention beyond literal meaning and should be conveyed adequately in translation. We propose to use lexical formality models to control the formality level of machine translation output. We demonstrate the effectiveness of our approach in empirical evaluations, as measured by automatic metrics and human assessments.

In a controlled experiment of sequence-to-sequence approaches for the task of sentence correction, we find that character-based models are generally more effective than word-based models and models that encode subword information via convolutions, and that modeling the output data as a series of diffs improves effectiveness over standard approaches. Our strongest sequence-to-sequence model improves over our strongest phrase-based statistical machine translation model, with access to the same data, by 6 M2 (0.5 GLEU) points. Additionally, in the data environment of the standard CoNLL-2014 setup, we demonstrate that modeling (and tuning against) diffs yields similar or better M2 scores with simpler models and/or significantly less data than previous sequence-to-sequence approaches.

We propose an approach to N-best list reranking using neural sequence-labelling models. We train a compositional model for error detection that calculates the probability of each token in a sentence being correct or incorrect, utilising the full sentence as context. Using the error detection model, we then re-rank the N best hypotheses generated by statistical machine translation systems. Our approach achieves state-of-the-art results on error correction for three different datasets, and it has the additional advantage of only using a small set of easily computed features that require no linguistic input.

In this paper, we model the problem of disfluency detection using a transition-based framework, which incrementally constructs and labels the disfluency chunk of input sentences using a new transition system without syntax information. Compared with sequence labeling methods, it can capture non-local chunk-level features; compared with joint parsing and disfluency detection methods, it is free for noise in syntax. Experiments show that our model achieves state-of-the-art f-score of 87.5% on the commonly used English Switchboard test set, and a set of in-house annotated Chinese data.

In this paper, we present a novel approach to infer significance of various textual edits to documents. An author may make several edits to a document; each edit varies in its impact to the content of the document. While some edits are surface changes and introduce negligible change, other edits may change the content/tone of the document significantly. In this paper, we perform an analysis on the human perceptions of edit importance while reviewing documents from one version to the next. We identify linguistic features that influence edit importance and model it in a regression based setting. We show that the predicted importance by our approach is highly correlated with the human perceived importance, established by a Mechanical Turk study.

Websites’ and mobile apps’ privacy policies, written in natural language, tend to be long and difficult to understand. Information privacy revolves around the fundamental principle of Notice and choice, namely the idea that users should be able to make informed decisions about what information about them can be collected and how it can be used. Internet users want control over their privacy, but their choices are often hidden in long and convoluted privacy policy texts. Moreover, little (if any) prior work has been done to detect the provision of choices in text. We address this challenge of enabling user choice by automatically identifying and extracting pertinent choice language in privacy policies. In particular, we present a two-stage architecture of classification models to identify opt-out choices in privacy policy text, labelling common varieties of choices with a mean F1 score of 0.735. Our techniques enable the creation of systems to help Internet users to learn about their choices, thereby effectuating notice and choice and improving Internet privacy.

Recent years have witnessed the proliferation of Massive Open Online Courses (MOOCs). With massive learners being offered MOOCs, there is a demand that the forum contents within MOOCs need to be classified in order to facilitate both learners and instructors. Therefore we investigate a significant application, which is to associate forum threads to subtitles of video clips. This task can be regarded as a document ranking problem, and the key is how to learn a distinguishable text representation from word sequences and learners’ behavior sequences. In this paper, we propose a novel cascade model, which can capture both the latent semantics and latent similarity by modeling MOOC data. Experimental results on two real-world datasets demonstrate that our textual representation outperforms state-of-the-art unsupervised counterparts for the application.

Explaining underlying causes or effects about events is a challenging but valuable task. We define a novel problem of generating explanations of a time series event by (1) searching cause and effect relationships of the time series with textual data and (2) constructing a connecting chain between them to generate an explanation. To detect causal features from text, we propose a novel method based on the Granger causality of time series between features extracted from text such as N-grams, topics, sentiments, and their composition. The generation of the sequence of causal entities requires a commonsense causative knowledge base with efficient reasoning. To ensure good interpretability and appropriate lexical usage we combine symbolic and neural representations, using a neural reasoning algorithm trained on commonsense causal tuples to predict the next cause step. Our quantitative and human analysis show empirical evidence that our method successfully extracts meaningful causality relationships between time series with textual features and generates appropriate explanation between them.

A document outlier is a document that substantially deviates in semantics from the majority ones in a corpus. Automatic identification of document outliers can be valuable in many applications, such as screening health records for medical mistakes. In this paper, we study the problem of mining semantically deviating document outliers in a given corpus. We develop a generative model to identify frequent and characteristic semantic regions in the word embedding space to represent the given corpus, and a robust outlierness measure which is resistant to noisy content in documents. Experiments conducted on two real-world textual data sets show that our method can achieve an up to 135% improvement over baselines in terms of recall at top-1% of the outlier ranking.

Duplicate documents are a pervasive problem in text datasets and can have a strong effect on unsupervised models. Methods to remove duplicate texts are typically heuristic or very expensive, so it is vital to know when and why they are needed. We measure the sensitivity of two latent semantic methods to the presence of different levels of document repetition. By artificially creating different forms of duplicate text we confirm several hypotheses about how repeated text impacts models. While a small amount of duplication is tolerable, substantial over-representation of subsets of the text may overwhelm meaningful topical patterns.

The charge prediction task is to determine appropriate charges for a given case, which is helpful for legal assistant systems where the user input is fact description. We argue that relevant law articles play an important role in this task, and therefore propose an attention-based neural network method to jointly model the charge prediction task and the relevant article extraction task in a unified framework. The experimental results show that, besides providing legal basis, the relevant articles can also clearly improve the charge prediction results, and our full model can effectively predict appropriate charges for cases with different expression styles.

One of the main obstacles for many Digital Humanities projects is the low data availability. Texts have to be digitized in an expensive and time consuming process whereas Optical Character Recognition (OCR) post-correction is one of the time-critical factors. At the example of OCR post-correction, we show the adaptation of a generic system to solve a specific problem with little data. The system accounts for a diversity of errors encountered in OCRed texts coming from different time periods in the domain of literature. We show that the combination of different approaches, such as e.g. Statistical Machine Translation and spell checking, with the help of a ranking mechanism tremendously improves over single-handed approaches. Since we consider the accessibility of the resulting tool as a crucial part of Digital Humanities collaborations, we describe the workflow we suggest for efficient text recognition and subsequent automatic and manual post-correction

Story detection is the task of determining whether or not a unit of text contains a story. Prior approaches achieved a maximum performance of 0.66 F1, and did not generalize well across different corpora. We present a new state-of-the-art detector that achieves a maximum performance of 0.75 F1 (a 14% improvement), with significantly greater generalizability than previous work. In particular, our detector achieves performance above 0.70 F1 across a variety of combinations of lexically different corpora for training and testing, as well as dramatic improvements (up to 4,000%) in performance when trained on a small, disfluent data set. The new detector uses two basic types of features–ones related to events, and ones related to characters–totaling 283 specific features overall; previous detectors used tens of thousands of features, and so this detector represents a significant simplification along with increased performance.

We present experiments that show the influence of native language on lexical choice when producing text in another language – in this particular case English. We start from the premise that non-native English speakers will choose lexical items that are close to words in their native language. This leads us to an etymology-based representation of documents written by people whose mother tongue is an Indo-European language. Based on this representation we grow a language family tree, that matches closely the Indo-European language tree.

Mining biomedical text offers an opportunity to automatically discover important facts and infer associations among them. As new scientific findings appear across a large collection of biomedical publications, our aim is to tap into this literature to automate biomedical knowledge extraction and identify important insights from them. Towards that goal, we develop a system with novel deep neural networks to extract insights on biomedical literature. Evaluation shows our system is able to provide insights with competitive accuracy of human acceptance and its relation extraction component outperforms previous work.

For accurate entity linking, we need to capture various information aspects of an entity, such as its description in a KB, contexts in which it is mentioned, and structured knowledge. Additionally, a linking system should work on texts from different domains without requiring domain-specific training data or hand-engineered features. In this work we present a neural, modular entity linking system that learns a unified dense representation for each entity using multiple sources of information, such as its description, contexts around its mentions, and its fine-grained types. We show that the resulting entity linking system is effective at combining these sources, and performs competitively, sometimes out-performing current state-of-the-art systems across datasets, without requiring any domain-specific training data or hand-engineered features. We also show that our model can effectively “embed” entities that are new to the KB, and is able to link its mentions accurately.