NAACL 2018 Papers

How to identify, extract, and use phrasal knowledge is a crucial problem for the task of Recognizing Textual Entailment (RTE). To solve this problem, we propose a method for detecting paraphrases via natural deduction proofs of semantic relations between sentence pairs. Our solution relies on a graph reformulation of partial variable unifications and an algorithm that induces subgraph alignments between meaning representations. Experiments show that our method can automatically detect various paraphrases that are absent from existing paraphrase databases. In addition, the detection of paraphrases using proof information improves the accuracy of RTE tasks.

Previous representation learning techniques for knowledge graph representation usually represent the same entity or relation in different triples with the same representation, without considering the ambiguity of relations and entities. To appropriately handle the semantic variety of entities/relations in distinct triples, we propose an accurate text-enhanced knowledge graph representation learning method, which can represent a relation/entity with different representations in different triples by exploiting additional textual information. Specifically, our method enhances representations by exploiting the entity descriptions and triple-specific relation mention. And a mutual attention mechanism between relation mention and entity description is proposed to learn more accurate textual representations for further improving knowledge graph representation. Experimental results show that our method achieves the state-of-the-art performance on both link prediction and triple classification tasks, and significantly outperforms previous text-enhanced knowledge representation models.

We present two neural models for event factuality prediction, which yield significant performance gains over previous models on three event factuality datasets: FactBank, UW, and MEANTIME. We also present a substantial expansion of the It Happened portion of the Universal Decompositional Semantics dataset, yielding the largest event factuality dataset to date. We report model results on this extended factuality dataset as well.

Traditional approaches to semantic parsing (SP) work by training individual models for each available parallel dataset of text-meaning pairs. In this paper, we explore the idea of polyglot semantic translation, or learning semantic parsing models that are trained on multiple datasets and natural languages. In particular, we focus on translating text to code signature representations using the software component datasets of Richardson and Kuhn (2017b,a). The advantage of such models is that they can be used for parsing a wide variety of input natural languages and output programming languages, or mixed input languages, using a single unified model. To facilitate modeling of this type, we develop a novel graph-based decoding framework that achieves state-of-the-art performance on the above datasets, and apply this method to two other benchmark SP tasks.

We present NEWSROOM, a summarization dataset of 1.3 million articles and summaries written by authors and editors in newsrooms of 38 major news publications. Extracted from search and social media metadata between 1998 and 2017, these high-quality summaries demonstrate high diversity of summarization styles. In particular, the summaries combine abstractive and extractive strategies, borrowing words and phrases from articles at varying rates. We analyze the extraction strategies used in NEWSROOM summaries against other datasets to quantify the diversity and difficulty of our new data, and train existing methods on the data to evaluate its utility and challenges. The dataset is available online at summari.es.

A major proportion of a text summary includes important entities found in the original text. These entities build up the topic of the summary. Moreover, they hold commonsense information once they are linked to a knowledge base. Based on these observations, this paper investigates the usage of linked entities to guide the decoder of a neural text summarizer to generate concise and better summaries. To this end, we leverage on an off-the-shelf entity linking system (ELS) to extract linked entities and propose Entity2Topic (E2T), a module easily attachable to a sequence-to-sequence model that transforms a list of entities into a vector representation of the topic of the summary. Current available ELS’s are still not sufficiently effective, possibly introducing unresolved ambiguities and irrelevant entities. We resolve the imperfections of the ELS by (a) encoding entities with selective disambiguation, and (b) pooling entity vectors using firm attention. By applying E2T to a simple sequenceto-sequence model with attention mechanism as base model, we see significant improvements of the performance in the Gigaword (sentence to title) and CNN (long document to multi-sentence highlights) summarization datasets by at least 2 ROUGE points.

Current measures for evaluating text simplification systems focus on evaluating lexical text aspects, neglecting its structural aspects. In this paper we propose the first measure to address structural aspects of text simplification, called SAMSA. It leverages recent advances in semantic parsing to assess simplification quality by decomposing the input based on its semantic structure and comparing it to the output. SAMSA provides a reference-less automatic evaluation procedure, avoiding the problems that reference-based methods face due to the vast space of valid simplifications for a given sentence. Our human evaluation experiments show both SAMSA’s substantial correlation with human judgments, as well as the deficiency of existing reference-based measures in evaluating structural simplification.

Word embeddings obtained from neural network models such as Word2Vec Skipgram have become popular representations of word meaning and have been evaluated on a variety of word similarity and relatedness norming data. Skipgram generates a set of word and context embeddings, the latter typically discarded after training. We demonstrate the usefulness of context embeddings in predicting asymmetric association between words from a recently published dataset of production norms (Jouravlev & McRae, 2016). Our findings suggest that humans respond with words closer to the cue within the context embedding space (rather than the word embedding space), when asked to generate thematically related words.

Temporal orientation refers to an individual’s tendency to connect to the psychological concepts of past, present or future, and it affects personality, motivation, emotion, decision making and stress coping processes. The study of the social media users’ psycho-demographic attributes from the perspective of human temporal orientation can be of utmost interest and importance to the business and administrative decision makers as it can provide an extra precious information for them to make informed decisions. In this paper, we propose a very first study to demonstrate the association between the sentiment view of the temporal orientation of the users and their different psycho-demographic attributes by analyzing their tweets. We first create a temporal orientation classifier in a minimally supervised way which classifies each tweet of the users in one of the three temporal categories, namely past, present, and future. A deep Bi-directional Long Short Term Memory (BLSTM) is used for the tweet classification task. Our tweet classifier achieves an accuracy of 78.27% when tested on a manually created test set. We then determine the users’ overall temporal orientation based on their tweets on the social media. The sentiment is added to the tweets at the fine-grained level where each temporal tweet is given a sentiment with either of the positive, negative or neutral. Our experiment reveals that depending upon the sentiment view of temporal orientation, a user’s attributes vary. We finally measure the correlation between the users’ sentiment view of temporal orientation and their different psycho-demographic factors using regression.

We introduce MeSys, a meaning-based approach, for solving English math word problems (MWPs) via understanding and reasoning in this paper. It first analyzes the text, transforms both body and question parts into their corresponding logic forms, and then performs inference on them. The associated context of each quantity is represented with proposed role-tags (e.g., nsubj, verb, etc.), which provides the flexibility for annotating an extracted math quantity with its associated context information (i.e., the physical meaning of this quantity). Statistical models are proposed to select the operator and operands. A noisy dataset is designed to assess if a solver solves MWPs mainly via understanding or mechanical pattern matching. Experimental results show that our approach outperforms existing systems on both benchmark datasets and the noisy dataset, which demonstrates that the proposed approach understands the meaning of each quantity in the text more.

Answering complex questions is a time-consuming activity for humans that requires reasoning and integration of information. Recent work on reading comprehension made headway in answering simple questions, but tackling complex questions is still an ongoing research challenge. Conversely, semantic parsers have been successful at handling compositionality, but only when the information resides in a target knowledge-base. In this paper, we present a novel framework for answering broad and complex questions, assuming answering simple questions is possible using a search engine and a reading comprehension model. We propose to decompose complex questions into a sequence of simple questions, and compute the final answer from the sequence of answers. To illustrate the viability of our approach, we create a new dataset of complex questions, ComplexWebQuestions, and present a model that decomposes questions and interacts with the web to compute an answer. We empirically demonstrate that question decomposition improves performance from 20.8 precision@1 to 27.5 precision@1 on this new dataset.

Building curious machines that can answer as well as ask questions is an important challenge for AI. The two tasks of question answering and question generation are usually tackled separately in the NLP literature. At the same time, both require significant amounts of supervised data which is hard to obtain in many domains. To alleviate these issues, we propose a self-training method for jointly learning to ask as well as answer questions, leveraging unlabeled text along with labeled question answer pairs for learning. We evaluate our approach on four benchmark datasets: SQUAD, MS MARCO, WikiQA and TrecQA, and show significant improvements over a number of established baselines on both question answering and question generation tasks. We also achieved new state-of-the-art results on two competitive answer sentence selection tasks: WikiQA and TrecQA.

Translation-based methods for grammar correction that directly map noisy, ungrammatical text to their clean counterparts are able to correct a broad range of errors; however, such techniques are bottlenecked by the need for a large parallel corpus of noisy and clean sentence pairs. In this paper, we consider synthesizing parallel data by noising a clean monolingual corpus. While most previous approaches introduce perturbations using features computed from local context windows, we instead develop error generation processes using a neural sequence transduction model trained to translate clean examples to their noisy counterparts. Given a corpus of clean examples, we propose beam search noising procedures to synthesize additional noisy examples that human evaluators were nearly unable to discriminate from nonsynthesized examples. Surprisingly, when trained on additional data synthesized using our best-performing noising scheme, our model approaches the same performance as when trained on additional nonsynthesized data.

Cross-lingual Hypernymy Detection involves determining if a word in one language (“fruit”) is a hypernym of a word in another language (“pomme” i.e. apple in French). The ability to detect hypernymy cross-lingually can aid in solving cross-lingual versions of tasks such as textual entailment and event coreference. We propose BiSparse-Dep, a family of unsupervised approaches for cross-lingual hypernymy detection, which learns sparse, bilingual word embeddings based on dependency contexts. We show that BiSparse-Dep can significantly improve performance on this task, compared to approaches based only on lexical context. Our approach is also robust, showing promise for low-resource settings: our dependency-based embeddings can be learned using a parser trained on related languages, with negligible loss in performance. We also crowd-source a challenging dataset for this task on four languages – Russian, French, Arabic, and Chinese. Our embeddings and datasets are publicly available.

Previously, neural methods in grammatical error correction (GEC) did not reach state-of-the-art results compared to phrase-based statistical machine translation (SMT) baselines. We demonstrate parallels between neural GEC and low-resource neural MT and successfully adapt several methods from low-resource MT to neural GEC. We further establish guidelines for trustable results in neural GEC and propose a set of model-independent methods for neural GEC that can be easily applied in most GEC settings. Proposed methods include adding source-side noise, domain-adaptation techniques, a GEC-specific training-objective, transfer learning with monolingual data, and ensembling of independently trained GEC models and language models. The combined effects of these methods result in better than state-of-the-art neural GEC models that outperform previously best neural GEC systems by more than 10% M² on the CoNLL-2014 benchmark and 5.9% on the JFLEG test set. Non-neural state-of-the-art systems are outperformed by more than 2% on the CoNLL-2014 benchmark and by 4% on JFLEG.

For over a decade, machine learning has been used to extract opinion-holder-target structures from text to answer the question “Who expressed what kind of sentiment towards what?”. Recent neural approaches do not outperform the state-of-the-art feature-based models for Opinion Role Labeling (ORL). We suspect this is due to the scarcity of labeled training data and address this issue using different multi-task learning (MTL) techniques with a related task which has substantially more data, i.e. Semantic Role Labeling (SRL). We show that two MTL models improve significantly over the single-task model for labeling of both holders and targets, on the development and the test sets. We found that the vanilla MTL model, which makes predictions using only shared ORL and SRL features, performs the best. With deeper analysis we determine what works and what might be done to make further improvements for ORL.

Efficient word representations play an important role in solving various problems related to Natural Language Processing (NLP), data mining, text mining etc. The issue of data sparsity poses a great challenge in creating efficient word representation model for solving the underlying problem. The problem is more intensified in resource-poor scenario due to the absence of sufficient amount of corpus. In this work we propose to minimize the effect of data sparsity by leveraging bilingual word embeddings learned through a parallel corpus. We train and evaluate Long Short Term Memory (LSTM) based architecture for aspect level sentiment classification. The neural network architecture is further assisted by the hand-crafted features for the prediction. We show the efficacy of the proposed model against state-of-the-art methods in two experimental setups i.e. multi-lingual and cross-lingual.

Text might contain or invoke multiple emotions with varying intensities. As such, emotion detection, to predict multiple emotions associated with a given text, can be cast into a multi-label classification problem. We would like to go one step further so that a ranked list of relevant emotions are generated where top ranked emotions are more intensely associated with text compared to lower ranked emotions, whereas the rankings of irrelevant emotions are not important. A novel framework of relevant emotion ranking is proposed to tackle the problem. In the framework, the objective loss function is designed elaborately so that both emotion prediction and rankings of only relevant emotions can be achieved. Moreover, we observe that some emotions co-occur more often while other emotions rarely co-exist. Such information is incorporated into the framework as constraints to improve the accuracy of emotion detection. Experimental results on two real-world corpora show that the proposed framework can effectively deal with emotion detection and performs remarkably better than the state-of-the-art emotion detection approaches and multi-label learning methods.

Target-dependent classification tasks, such as aspect-level sentiment analysis, perform fine-grained classifications towards specific targets. Semantic compositions over tree structures are promising for such tasks, as they can potentially capture long-distance interactions between targets and their contexts. However, previous work that operates on tree structures resorts to syntactic parsers or Treebank annotations, which are either subject to noise in informal texts or highly expensive to obtain. To address above issues, we propose a reinforcement learning based approach, which automatically induces target-specific sentence representations over tree structures. The underlying model is a RNN encoder-decoder that explores possible binary tree structures and a reward mechanism that encourages structures that improve performances on downstream tasks. We evaluate our approach on two benchmark tasks: firm-specific cumulative abnormal return prediction (based on formal news texts) and aspect-level sentiment analysis (based on informal social media texts). Experimental results show that our model gives superior performances compared to previous work that operates on parsed trees. Moreover, our approach gives some intuitions on how target-specific sentence representations can be achieved from its word constituents.