COLING 2018 论文列表

Approaches to authorship attribution, the task of identifying the author of a document, are based on analysis of individuals’ writing style and/or preferred topics. Although the problem has been widely explored, no previous studies have analysed the relationship between dataset characteristics and effectiveness of different types of features. This study carries out an analysis of four widely used datasets to explore how different types of features affect authorship attribution accuracy under varying conditions. The results of the analysis are applied to authorship attribution models based on both discrete and continuous representations. We apply the conclusions from our analysis to an extension of an existing approach to authorship attribution and outperform the prior state-of-the-art on two out of the four datasets used.

Learning from a real-world data stream and continuously updating the model without explicit supervision is a new challenge for NLP applications with machine learning components. In this work, we have developed an adaptive learning system for text simplification, which improves the underlying learning-to-rank model from usage data, i.e. how users have employed the system for the task of simplification. Our experimental result shows that, over a period of time, the performance of the embedded paraphrase ranking model increases steadily improving from a score of 62.88% up to 75.70% based on the NDCG@10 evaluation metrics. To our knowledge, this is the first study where an NLP component is adaptively improved through usage.

Complexity of texts is usually assessed only at the lexical and syntactic levels. Although it is known that conceptual complexity plays a significant role in text understanding, no attempts have been made at assessing it automatically. We propose to automatically estimate the conceptual complexity of texts by exploiting a number of graph-based measures on a large knowledge base. By using a high-quality language learners corpus for English, we show that graph-based measures of individual text concepts, as well as the way they relate to each other in the knowledge graph, have a high discriminative power when distinguishing between two versions of the same text. Furthermore, when used as features in a binary classification task aiming to choose the simpler of two versions of the same text, our measures achieve high performance even in a default setup.

We analyze two novel data sets of German educational media texts targeting adults and children. The analysis is based on 400 automatically extracted measures of linguistic complexity from a wide range of linguistic domains. We show that both data sets exhibit broad linguistic adaptation to the target audience, which generalizes across both data sets. Our most successful binary classification model for German readability robustly shows high accuracy between 89.4%–98.9% for both data sets. To our knowledge, this comprehensive German readability model is the first for which robust cross-corpus performance has been shown. The research also contributes resources for German readability assessment that are externally validated as successful for different target audiences: we compiled a new corpus of German news broadcast subtitles, the Tagesschau/Logo corpus, and crawled a GEO/GEOlino corpus substantially enlarging the data compiled by Hancke et al. 2012.

Recently, a significant number of studies have focused on neural information retrieval (IR) models. One category of works use unlabeled data to train general word embeddings based on term proximity, which can be integrated into traditional IR models. The other category employs labeled data (e.g. click-through data) to train end-to-end neural IR models consisting of layers for target-specific representation learning. The latter idea accounts better for the IR task and is favored by recent research works, which is the one we will follow in this paper. We hypothesize that general semantics learned from unlabeled data can complement task-specific representation learned from labeled data of limited quality, and that a combination of the two is favorable. To this end, we propose a learning framework which can benefit from both labeled and more abundant unlabeled data for representation learning in the context of IR. Through a joint learning fashion in a single neural framework, the learned representation is optimized to minimize both the supervised loss on query-document matching and the unsupervised loss on text reconstruction. Standard retrieval experiments on TREC collections indicate that the joint learning methodology leads to significant better performance of retrieval over several strong baselines for IR.

Fine-grained entity typing aims at identifying the semantic type of an entity in KB. Type information is very important in knowledge bases, but are unfortunately incomplete even in some large knowledge bases. Limitations of existing methods are either ignoring the structure and type information in KB or requiring large scale annotated corpus. To address these issues, we propose an attributed and predictive entity embedding method, which can fully utilize various kinds of information comprehensively. Extensive experiments on two real DBpedia datasets show that our proposed method significantly outperforms 8 state-of-the-art methods, with 4.0% and 5.2% improvement in Mi-F1 and Ma-F1, respectively.

In this paper, we study how we can improve a deep learning approach to textual entailment by incorporating lexical entailment relations from WordNet. Our idea is to embed the lexical entailment knowledge contained in WordNet in specially-learned word vectors, which we call “entailment vectors.” We present a standard neural network model and a novel set-theoretic model to learn these entailment vectors from word pairs with known lexical entailment relations derived from WordNet. We further incorporate these entailment vectors into a decomposable attention model for textual entailment and evaluate the model on the SICK and the SNLI dataset. We find that using these special entailment word vectors, we can significantly improve the performance of textual entailment compared with a baseline that uses only standard word2vec vectors. The final performance of our model is close to or above the state of the art, but our method does not rely on any manually-crafted rules or extensive syntactic features.

Identifying emergent research trends is a key issue for both primary researchers as well as secondary research managers. Such processes can uncover the historical development of an area, and yield insight on developing topics. We propose an embedded trend detection framework for this task which incorporates our bijunctive hypothesis that important phrases are written by important authors within a field and vice versa. By ranking both author and phrase information in a multigraph, our method jointly determines key phrases and authoritative authors. We represent this intermediate output as phrasal embeddings, and feed this to a recurrent neural network (RNN) to compute trend scores that identify research trends. Over two large datasets of scientific articles, we demonstrate that our approach successfully detects past trends from the field, outperforming baselines based solely on text centrality or citation.

Most previous research in text simplification has aimed to develop generic solutions, assuming very homogeneous target audiences with consistent intra-group simplification needs. We argue that this assumption does not hold, and that instead we need to develop simplification systems that adapt to the individual needs of specific users. As a first step towards personalized simplification, we propose a framework for adaptive lexical simplification and introduce Lexi, a free open-source and easily extensible tool for adaptive, personalized text simplification. Lexi is easily installed as a browser extension, enabling easy access to the service for its users.

Distributional word representations (often referred to as word embeddings) are omnipresent in modern NLP. Early work has focused on building representations for word types, and recent studies show that lemmatization and part of speech (POS) disambiguation of targets in isolation improve the performance of word embeddings on a range of downstream tasks. However, the reasons behind these improvements, the qualitative effects of these operations and the combined performance of lemmatized and POS disambiguated targets are less studied. This work aims to close this gap and puts previous findings into a general perspective. We examine the effect of lemmatization and POS typing on word embedding performance in a novel resource-based evaluation scenario, as well as on standard similarity benchmarks. We show that these two operations have complimentary qualitative and vocabulary-level effects and are best used in combination. We find that the improvement is more pronounced for verbs and show how lemmatization and POS typing implicitly target some of the verb-specific issues. We claim that the observed improvement is a result of better conceptual alignment between word embeddings and lexical resources, stressing the need for conceptually plausible modeling of word embedding targets.

A lexical simplification (LS) system aims to substitute complex words with simple words in a text, while preserving its meaning and grammaticality. Despite individual users’ differences in vocabulary knowledge, current systems do not consider these variations; rather, they are trained to find one optimal substitution or ranked list of substitutions for all users. We evaluate the performance of a state-of-the-art LS system on individual learners of English at different proficiency levels, and measure the benefits of using complex word identification (CWI) models to personalize the system. Experimental results show that even a simple personalized CWI model, based on graded vocabulary lists, can help the system avoid some unnecessary simplifications and produce more readable output.

Sequence-to-sequence model has been applied to solve math word problems. The model takes math problem descriptions as input and generates equations as output. The advantage of sequence-to-sequence model requires no feature engineering and can generate equations that do not exist in training data. However, our experimental analysis reveals that this model suffers from two shortcomings: (1) generate spurious numbers; (2) generate numbers at wrong positions. In this paper, we propose incorporating copy and alignment mechanism to the sequence-to-sequence model (namely CASS) to address these shortcomings. To train our model, we apply reinforcement learning to directly optimize the solution accuracy. It overcomes the “train-test discrepancy” issue of maximum likelihood estimation, which uses the surrogate objective of maximizing equation likelihood during training while the evaluation metric is solution accuracy (non-differentiable) at test time. Furthermore, to explore the effectiveness of our neural model, we use our model output as a feature and incorporate it into the feature-based model. Experimental results show that (1) The copy and alignment mechanism is effective to address the two issues; (2) Reinforcement learning leads to better performance than maximum likelihood on this task; (3) Our neural model is complementary to the feature-based model and their combination significantly outperforms the state-of-the-art results.

We present domain independent models to date documents based only on neologism usage patterns. Our models capture patterns of neologism usage over time to date texts, provide insights into temporal locality of word usage over a span of 150 years, and generalize to various domains like News, Fiction, and Non-Fiction with competitive performance. Quite intriguingly, we show that by modeling only the distribution of usage counts over neologisms (the model being agnostic of the particular words themselves), we achieve competitive performance using several orders of magnitude fewer features (only 200 input features) compared to state of the art models some of which use 200K features.

In this paper, we describe TextEnt, a neural network model that learns distributed representations of entities and documents directly from a knowledge base (KB). Given a document in a KB consisting of words and entity annotations, we train our model to predict the entity that the document describes and map the document and its target entity close to each other in a continuous vector space. Our model is trained using a large number of documents extracted from Wikipedia. The performance of the proposed model is evaluated using two tasks, namely fine-grained entity typing and multiclass text classification. The results demonstrate that our model achieves state-of-the-art performance on both tasks. The code and the trained representations are made available online for further academic research.

Event relation recognition is a challenging language processing task. It is required to determine the relation class of a pair of query events, such as causality, under the condition that there isn’t any reliable clue for use. We follow the traditional statistical approach in this paper, speculating the relation class of the target events based on the relation-class distributions on the similar events. There is minimal supervision used during the speculation process. In particular, we incorporate image processing into the acquisition of similar event instances, including the utilization of images for visually representing event scenes, and the use of the neural network based image matching for approximate calculation between events. We test our method on the ACE-R2 corpus and compared our model with the fully-supervised neural network models. Experimental results show that we achieve a comparable performance to CNN while slightly better than LSTM.

Reading comprehension models are based on recurrent neural networks that sequentially process the document tokens. As interest turns to answering more complex questions over longer documents, sequential reading of large portions of text becomes a substantial bottleneck. Inspired by how humans use document structure, we propose a novel framework for reading comprehension. We represent documents as trees, and model an agent that learns to interleave quick navigation through the document tree with more expensive answer extraction. To encourage exploration of the document tree, we propose a new algorithm, based on Deep Q-Network (DQN), which strategically samples tree nodes at training time. Empirically we find our algorithm improves question answering performance compared to DQN and a strong information-retrieval (IR) baseline, and that ensembling our model with the IR baseline results in further gains in performance.

Modeling U.S. Congressional legislation and roll-call votes has received significant attention in previous literature, and while legislators across 50 state governments and D.C. propose over 100,000 bills each year, enacting over 30% of them on average, state level analysis has received relatively less attention due in part to the difficulty in obtaining the necessary data. Since each state legislature is guided by their own procedures, politics and issues, however, it is difficult to qualitatively asses the factors that affect the likelihood of a legislative initiative succeeding. We present several methods for modeling the likelihood of a bill receiving floor action across all 50 states and D.C. We utilize the lexical content of over 1 million bills, along with contextual legislature and legislator derived features to build our predictive models, allowing a comparison of what factors are important to the lawmaking process. Furthermore, we show that these signals hold complementary predictive power, together achieving an average improvement in accuracy of 18% over state specific baselines.

The paper explores the ability of LSTM networks trained on a language modeling task to detect linguistic structures which are ungrammatical due to extraction violations (extra arguments and subject-relative clause island violations), and considers its implications for the debate on language innatism. The results show that the current RNN model can correctly classify (un)grammatical sentences, in certain conditions, but it is sensitive to linguistic processing factors and probably ultimately unable to induce a more abstract notion of grammaticality, at least in the domain we tested.

Non-local features have been exploited by syntactic parsers for capturing dependencies between sub output structures. Such features have been a key to the success of state-of-the-art statistical parsers. With the rise of deep learning, however, it has been shown that local output decisions can give highly competitive accuracies, thanks to the power of dense neural input representations that embody global syntactic information. We investigate two conceptually simple local neural models for constituent parsing, which make local decisions to constituent spans and CFG rules, respectively. Consistent with previous findings along the line, our best model gives highly competitive results, achieving the labeled bracketing F1 scores of 92.4% on PTB and 87.3% on CTB 5.1.