I explore the hypothesis that conventional neural network models (e.g., recurrent neural networks) are incorrectly biased for making linguistically sensible generalizations when learning, and that a better class of models is based on architectures that reflect hierarchical structures for which considerable behavioral evidence exists. I focus on the problem of modeling and representing the meanings of sentences. On the generation front, I introduce recurrent neural network grammars (RNNGs), a joint, generative model of phrase-structure trees and sentences. RNNGs operate via a recursive syntactic process reminiscent of probabilistic context-free grammar generation, but decisions are parameterized using RNNs that condition on the entire (top-down, left-to-right) syntactic derivation history, thus relaxing context-free independence assumptions, while retaining a bias toward explaining decisions via “syntactically local” conditioning contexts. Experiments show that RNNGs obtain better results in generating language than models that don’t exploit linguistic structure. On the representation front, I explore unsupervised learning of syntactic structures based on distant semantic supervision using a reinforcement-learning algorithm. The learner seeks a syntactic structure that provides a compositional architecture that produces a good representation for a downstream semantic task. Although the inferred structures are quite different from traditional syntactic analyses, the performance on the downstream tasks surpasses that of systems that use sequential RNNs and tree-structured RNNs based on treebank dependencies. This is joint work with Adhi Kuncoro, Dani Yogatama, Miguel Ballesteros, Phil Blunsom, Ed Grefenstette, Wang Ling, and Noah A. Smith.

Language acquisition can be modeled as a statistical inference problem: children use sentences and sounds in their input to infer linguistic structure. However, in many cases, children learn from data whose statistical structure is distorted relative to the language they are learning. Such distortions can arise either in the input itself, or as a result of children’s immature strategies for encoding their input. This work examines several cases in which the statistical structure of children’s input differs from the language being learned. Analyses show that these distortions of the input can be accounted for with a statistical learning framework by carefully considering the inference problems that learners solve during language acquisition

Recent work has explored the syntactic abilities of RNNs using the subject-verb agreement task, which diagnoses sensitivity to sentence structure. RNNs performed this task well in common cases, but faltered in complex sentences (Linzen et al., 2016). We test whether these errors are due to inherent limitations of the architecture or to the relatively indirect supervision provided by most agreement dependencies in a corpus. We trained a single RNN to perform both the agreement task and an additional task, either CCG supertagging or language modeling. Multi-task training led to significantly lower error rates, in particular on complex sentences, suggesting that RNNs have the ability to evolve more sophisticated syntactic representations than shown before. We also show that easily available agreement training data can improve performance on other syntactic tasks, in particular when only a limited amount of training data is available for those tasks. The multi-task paradigm can also be leveraged to inject grammatical knowledge into language models.

A writer’s style depends not just on personal traits but also on her intent and mental state. In this paper, we show how variants of the same writing task can lead to measurable differences in writing style. We present a case study based on the story cloze task (Mostafazadeh et al., 2016a), where annotators were assigned similar writing tasks with different constraints: (1) writing an entire story, (2) adding a story ending for a given story context, and (3) adding an incoherent ending to a story. We show that a simple linear classifier informed by stylistic features is able to successfully distinguish among the three cases, without even looking at the story context. In addition, combining our stylistic features with language model predictions reaches state of the art performance on the story cloze challenge. Our results demonstrate that different task framings can dramatically affect the way people write.

This paper is concerned with building deep grammatical relation (GR) analysis using data-driven approach. To deal with this problem, we propose graph merging, a new perspective, for building flexible dependency graphs: Constructing complex graphs via constructing simple subgraphs. We discuss two key problems in this perspective: (1) how to decompose a complex graph into simple subgraphs, and (2) how to combine subgraphs into a coherent complex graph. Experiments demonstrate the effectiveness of graph merging. Our parser reaches state-of-the-art performance and is significantly better than two transition-based parsers.

Metaphor detection has been both challenging and rewarding in natural language processing applications. This study offers a new approach based on eventive information in detecting metaphors by leveraging the Chinese writing system, which is a culturally bound ontological system organized according to the basic concepts represented by radicals. As such, the information represented is available in all Chinese text without pre-processing. Since metaphor detection is another culturally based conceptual representation, we hypothesize that sub-textual information can facilitate the identification and classification of the types of metaphoric events denoted in Chinese text. We propose a set of syntactic conditions crucial to event structures to improve the model based on the classification of radical groups. With the proposed syntactic conditions, the model achieves a performance of 0.8859 in terms of F-scores, making 1.7% of improvement than the same classifier with only Bag-of-word features. Results show that eventive information can improve the effectiveness of metaphor detection. Event information is rooted in every language, and thus this approach has a high potential to be applied to metaphor detection in other languages.

This paper presents a collaborative partitioning algorithm—a novel ensemble-based approach to coreference resolution. Starting from the all-singleton partition, we search for a solution close to the ensemble’s outputs in terms of a task-specific similarity measure. Our approach assumes a loose integration of individual components of the ensemble and can therefore combine arbitrary coreference resolvers, regardless of their models. Our experiments on the CoNLL dataset show that collaborative partitioning yields results superior to those attained by the individual components, for ensembles of both strong and weak systems. Moreover, by applying the collaborative partitioning algorithm on top of three state-of-the-art resolvers, we obtain the best coreference performance reported so far in the literature (MELA v08 score of 64.47).

We address the task of Named Entity Disambiguation (NED) for noisy text. We present WikilinksNED, a large-scale NED dataset of text fragments from the web, which is significantly noisier and more challenging than existing news-based datasets. To capture the limited and noisy local context surrounding each mention, we design a neural model and train it with a novel method for sampling informative negative examples. We also describe a new way of initializing word and entity embeddings that significantly improves performance. Our model significantly outperforms existing state-of-the-art methods on WikilinksNED while achieving comparable performance on a smaller newswire dataset.

For many applications of question answering (QA), being able to explain why a given model chose an answer is critical. However, the lack of labeled data for answer justifications makes learning this difficult and expensive. Here we propose an approach that uses answer ranking as distant supervision for learning how to select informative justifications, where justifications serve as inferential connections between the question and the correct answer while often containing little lexical overlap with either. We propose a neural network architecture for QA that reranks answer justifications as an intermediate (and human-interpretable) step in answer selection. Our approach is informed by a set of features designed to combine both learned representations and explicit features to capture the connection between questions, answers, and answer justifications. We show that with this end-to-end approach we are able to significantly improve upon a strong IR baseline in both justification ranking (+9% rated highly relevant) and answer selection (+6% P@1).

Question answering (QA) systems are easily distracted by irrelevant or redundant words in questions, especially when faced with long or multi-sentence questions in difficult domains. This paper introduces and studies the notion of essential question terms with the goal of improving such QA solvers. We illustrate the importance of essential question terms by showing that humans’ ability to answer questions drops significantly when essential terms are eliminated from questions. We then develop a classifier that reliably (90% mean average precision) identifies and ranks essential terms in questions. Finally, we use the classifier to demonstrate that the notion of question term essentiality allows state-of-the-art QA solver for elementary-level science questions to make better and more informed decisions,improving performance by up to 5%.We also introduce a new dataset of over 2,200 crowd-sourced essential terms annotated science questions.

Pairwise ranking methods are the most widely used discriminative training approaches for structure prediction problems in natural language processing (NLP). Decomposing the problem of ranking hypotheses into pairwise comparisons enables simple and efficient solutions. However, neglecting the global ordering of the hypothesis list may hinder learning. We propose a listwise learning framework for structure prediction problems such as machine translation. Our framework directly models the entire translation list’s ordering to learn parameters which may better fit the given listwise samples. Furthermore, we propose top-rank enhanced loss functions, which are more sensitive to ranking errors at higher positions. Experiments on a large-scale Chinese-English translation task show that both our listwise learning framework and top-rank enhanced listwise losses lead to significant improvements in translation quality.

Word embeddings are widely used in Natural Language Processing, mainly due to their success in capturing semantic information from massive corpora. However, their creation process does not allow the different meanings of a word to be automatically separated, as it conflates them into a single vector. We address this issue by proposing a new model which learns word and sense embeddings jointly. Our model exploits large corpora and knowledge from semantic networks in order to produce a unified vector space of word and sense embeddings. We evaluate the main features of our approach both qualitatively and quantitatively in a variety of tasks, highlighting the advantages of the proposed method in comparison to state-of-the-art word- and sense-based models.

This paper is concerned with identifying contexts useful for training word representation models for different word classes such as adjectives (A), verbs (V), and nouns (N). We introduce a simple yet effective framework for an automatic selection of class-specific context configurations. We construct a context configuration space based on universal dependency relations between words, and efficiently search this space with an adapted beam search algorithm. In word similarity tasks for each word class, we show that our framework is both effective and efficient. Particularly, it improves the Spearman’s rho correlation with human scores on SimLex-999 over the best previously proposed class-specific contexts by 6 (A), 6 (V) and 5 (N) rho points. With our selected context configurations, we train on only 14% (A), 26.2% (V), and 33.6% (N) of all dependency-based contexts, resulting in a reduced training time. Our results generalise: we show that the configurations our algorithm learns for one English training setup outperform previously proposed context types in another training setup for English. Moreover, basing the configuration space on universal dependencies, it is possible to transfer the learned configurations to German and Italian. We also demonstrate improved per-class results over other context types in these two languages..

Vector representations of word meaning have found many applications in the field of natural language processing. Word vectors intuitively represent the average context in which a given word tends to occur, but they cannot explicitly model the diversity of these contexts. Although region representations of word meaning offer a natural alternative to word vectors, only few methods have been proposed that can effectively learn word regions. In this paper, we propose a new word embedding model which is based on SVM regression. We show that the underlying ranking interpretation of word contexts is sufficient to match, and sometimes outperform, the performance of popular methods such as Skip-gram. Furthermore, we show that by using a quadratic kernel, we can effectively learn word regions, which outperform existing unsupervised models for the task of hypernym detection.

Recent studies of distributional semantic models have set up a competition between word embeddings obtained from predictive neural networks and word vectors obtained from abstractive count-based models. This paper is an attempt to reveal the underlying contribution of additional training data and post-processing steps on each type of model in word similarity and relatedness inference tasks. We do so by designing an artificial language framework, training a predictive and a count-based model on data sampled from this grammar, and evaluating the resulting word vectors in paradigmatic and syntagmatic tasks defined with respect to the grammar.

Conventional word embeddings are trained with specific criteria (e.g., based on language modeling or co-occurrence) inside a single information source, disregarding the opportunity for further calibration using external knowledge. This paper presents a unified framework that leverages pre-learned or external priors, in the form of a regularizer, for enhancing conventional language model-based embedding learning. We consider two types of regularizers. The first type is derived from topic distribution by running LDA on unlabeled data. The second type is based on dictionaries that are created with human annotation efforts. To effectively learn with the regularizers, we propose a novel data structure, trajectory softmax, in this paper. The resulting embeddings are evaluated by word similarity and sentiment classification. Experimental results show that our learning framework with regularization from prior knowledge improves embedding quality across multiple datasets, compared to a diverse collection of baseline methods.

Neural network models have recently been applied to the task of automatic essay scoring, giving promising results. Existing work used recurrent neural networks and convolutional neural networks to model input essays, giving grades based on a single vector representation of the essay. On the other hand, the relative advantages of RNNs and CNNs have not been compared. In addition, different parts of the essay can contribute differently for scoring, which is not captured by existing models. We address these issues by building a hierarchical sentence-document model to represent essays, using the attention mechanism to automatically decide the relative weights of words and sentences. Results show that our model outperforms the previous state-of-the-art methods, demonstrating the effectiveness of the attention mechanism.

This paper proposes a matching technique for learning causal associations between word features and class labels in document classification. The goal is to identify more meaningful and generalizable features than with only correlational approaches. Experiments with sentiment classification show that the proposed method identifies interpretable word associations with sentiment and improves classification performance in a majority of cases. The proposed feature selection method is particularly effective when applied to out-of-domain data.

Understanding stories – sequences of events – is a crucial yet challenging natural language understanding task. These events typically carry multiple aspects of semantics including actions, entities and emotions. Not only does each individual aspect contribute to the meaning of the story, so does the interaction among these aspects. Building on this intuition, we propose to jointly model important aspects of semantic knowledge – frames, entities and sentiments – via a semantic language model. We achieve this by first representing these aspects’ semantic units at an appropriate level of abstraction and then using the resulting vector representations for each semantic aspect to learn a joint representation via a neural language model. We show that the joint semantic language model is of high quality and can generate better semantic sequences than models that operate on the word level. We further demonstrate that our joint model can be applied to story cloze test and shallow discourse parsing tasks with improved performance and that each semantic aspect contributes to the model.

Learning internal word structure has recently been recognized as an important step in various multilingual processing tasks and in theoretical language comparison. In this paper, we present a neural encoder-decoder model for learning canonical morphological segmentation. Our model combines character-level sequence-to-sequence transformation with a language model over canonical segments. We obtain up to 4% improvement over a strong character-level encoder-decoder baseline for three languages. Our model outperforms the previous state-of-the-art for two languages, while eliminating the need for external resources such as large dictionaries. Finally, by comparing the performance of encoder-decoder and classical statistical machine translation systems trained with and without corpus counts, we show that including corpus counts is beneficial to both approaches.

Automatic summarisation is a popular approach to reduce a document to its main arguments. Recent research in the area has focused on neural approaches to summarisation, which can be very data-hungry. However, few large datasets exist and none for the traditionally popular domain of scientific publications, which opens up challenging research avenues centered on encoding large, complex documents. In this paper, we introduce a new dataset for summarisation of computer science publications by exploiting a large resource of author provided summaries and show straightforward ways of extending it further. We develop models on the dataset making use of both neural sentence encoding and traditionally used summarisation features and show that models which encode sentences as well as their local and global context perform best, significantly outperforming well-established baseline methods.

Topic models jointly learn topics and document-level topic distribution. Extrinsic evaluation of topic models tends to focus exclusively on topic-level evaluation, e.g. by assessing the coherence of topics. We demonstrate that there can be large discrepancies between topic- and document-level model quality, and that basing model evaluation on topic-level analysis can be highly misleading. We propose a method for automatically predicting topic model quality based on analysis of document-level topic allocations, and provide empirical evidence for its robustness.

This paper presents a novel approach to character identification, that is an entity linking task that maps mentions to characters in dialogues from TV show transcripts. We first augment and correct several cases of annotation errors in an existing corpus so the corpus is clearer and cleaner for statistical learning. We also introduce the agglomerative convolutional neural network that takes groups of features and learns mention and mention-pair embeddings for coreference resolution. We then propose another neural model that employs the embeddings learned and creates cluster embeddings for entity linking. Our coreference resolution model shows comparable results to other state-of-the-art systems. Our entity linking model significantly outperforms the previous work, showing the F1 score of 86.76% and the accuracy of 95.30% for character identification.

We address the problem of cross-language adaptation for question-question similarity reranking in community question answering, with the objective to port a system trained on one input language to another input language given labeled training data for the first language and only unlabeled data for the second language. In particular, we propose to use adversarial training of neural networks to learn high-level features that are discriminative for the main learning task, and at the same time are invariant across the input languages. The evaluation results show sizable improvements for our cross-language adversarial neural network (CLANN) model over a strong non-adversarial system.

We present a feature-rich knowledge tracing method that captures a student’s acquisition and retention of knowledge during a foreign language phrase learning task. We model the student’s behavior as making predictions under a log-linear model, and adopt a neural gating mechanism to model how the student updates their log-linear parameters in response to feedback. The gating mechanism allows the model to learn complex patterns of retention and acquisition for each feature, while the log-linear parameterization results in an interpretable knowledge state. We collect human data and evaluate several versions of the model.

We present a generative model of natural language sentences and demonstrate its application to semantic parsing. In the generative process, a logical form sampled from a prior, and conditioned on this logical form, a grammar probabilistically generates the output sentence. Grammar induction using MCMC is applied to learn the grammar given a set of labeled sentences with corresponding logical forms. We develop a semantic parser that finds the logical form with the highest posterior probability exactly. We obtain strong results on the GeoQuery dataset and achieve state-of-the-art F1 on Jobs.

Tree kernels (TKs) and neural networks are two effective approaches for automatic feature engineering. In this paper, we combine them by modeling context word similarity in semantic TKs. This way, the latter can operate subtree matching by applying neural-based similarity on tree lexical nodes. We study how to learn representations for the words in context such that TKs can exploit more focused information. We found that neural embeddings produced by current methods do not provide a suitable contextual similarity. Thus, we define a new approach based on a Siamese Network, which produces word representations while learning a binary text similarity. We set the latter considering examples in the same category as similar. The experiments on question and sentiment classification show that our semantic TK highly improves previous results.

Recent development of large-scale question answering (QA) datasets triggered a substantial amount of research into end-to-end neural architectures for QA. Increasingly complex systems have been conceived without comparison to simpler neural baseline systems that would justify their complexity. In this work, we propose a simple heuristic that guides the development of neural baseline systems for the extractive QA task. We find that there are two ingredients necessary for building a high-performing neural QA system: first, the awareness of question words while processing the context and second, a composition function that goes beyond simple bag-of-words modeling, such as recurrent neural networks. Our results show that FastQA, a system that meets these two requirements, can achieve very competitive performance compared with existing models. We argue that this surprising finding puts results of previous systems and the complexity of recent QA datasets into perspective.

Factoid question answering (QA) has recently benefited from the development of deep learning (DL) systems. Neural network models outperform traditional approaches in domains where large datasets exist, such as SQuAD (ca. 100,000 questions) for Wikipedia articles. However, these systems have not yet been applied to QA in more specific domains, such as biomedicine, because datasets are generally too small to train a DL system from scratch. For example, the BioASQ dataset for biomedical QA comprises less then 900 factoid (single answer) and list (multiple answers) QA instances. In this work, we adapt a neural QA system trained on a large open-domain dataset (SQuAD, source) to a biomedical dataset (BioASQ, target) by employing various transfer learning techniques. Our network architecture is based on a state-of-the-art QA system, extended with biomedical word embeddings and a novel mechanism to answer list questions. In contrast to existing biomedical QA systems, our system does not rely on domain-specific ontologies, parsers or entity taggers, which are expensive to create. Despite this fact, our systems achieve state-of-the-art results on factoid questions and competitive results on list questions.

This paper explores a divisive hierarchical clustering algorithm based on the well-known Obligatory Contour Principle in phonology. The purpose is twofold: to see if such an algorithm could be used for unsupervised classification of phonemes or graphemes in corpora, and to investigate whether this purported universal constraint really holds for several classes of phonological distinctive features. The algorithm achieves very high accuracies in an unsupervised setting of inferring a consonant-vowel distinction, and also has a strong tendency to detect coronal phonemes in an unsupervised fashion. Remaining classes, however, do not correspond as neatly to phonological distinctive feature splits. While the results offer only mixed support for a universal Obligatory Contour Principle, the algorithm can be very useful for many NLP tasks due to the high accuracy in revealing consonant/vowel/coronal distinctions.

Previous studies have shown that investor sentiment indicators can predict stock market change. A domain-specific sentiment lexicon and sentiment-oriented word embedding model would help the sentiment analysis in financial domain and stock market. In this paper, we present a new approach to learning stock market lexicon from StockTwits, a popular financial social network for investors to share ideas. It learns word polarity by predicting message sentiment, using a neural net-work. The sentiment-oriented word embeddings are learned from tens of millions of StockTwits posts, and this is the first study presenting sentiment-oriented word embeddings for stock market. The experiments of predicting investor sentiment show that our lexicon outperformed other lexicons built by the state-of-the-art methods, and the sentiment-oriented word vector was much better than the general word embeddings.

The task of relation classification in the biomedical domain is complex due to the presence of samples obtained from heterogeneous sources such as research articles, discharge summaries, or electronic health records. It is also a constraint for classifiers which employ manual feature engineering. In this paper, we propose a convolutional recurrent neural network (CRNN) architecture that combines RNNs and CNNs in sequence to solve this problem. The rationale behind our approach is that CNNs can effectively identify coarse-grained local features in a sentence, while RNNs are more suited for long-term dependencies. We compare our CRNN model with several baselines on two biomedical datasets, namely the i2b2-2010 clinical relation extraction challenge dataset, and the SemEval-2013 DDI extraction dataset. We also evaluate an attentive pooling technique and report its performance in comparison with the conventional max pooling method. Our results indicate that the proposed model achieves state-of-the-art performance on both datasets.

Idea Density (ID) measures the rate at which ideas or elementary predications are expressed in an utterance or in a text. Lower ID is found to be associated with an increased risk of developing Alzheimer’s disease (AD) (Snowdon et al., 1996; Engelman et al., 2010). ID has been used in two different versions: propositional idea density (PID) counts the expressed ideas and can be applied to any text while semantic idea density (SID) counts pre-defined information content units and is naturally more applicable to normative domains, such as picture description tasks. In this paper, we develop DEPID, a novel dependency-based method for computing PID, and its version DEPID-R that enables to exclude repeating ideas—a feature characteristic to AD speech. We conduct the first comparison of automatically extracted PID and SID in the diagnostic classification task on two different AD datasets covering both closed-topic and free-recall domains. While SID performs better on the normative dataset, adding PID leads to a small but significant improvement (+1.7 F-score). On the free-topic dataset, PID performs better than SID as expected (77.6 vs 72.3 in F-score) but adding the features derived from the word embedding clustering underlying the automatic SID increases the results considerably, leading to an F-score of 84.8.

We show that relation extraction can be reduced to answering simple reading comprehension questions, by associating one or more natural-language questions with each relation slot. This reduction has several advantages: we can (1) learn relation-extraction models by extending recent neural reading-comprehension techniques, (2) build very large training sets for those models by combining relation-specific crowd-sourced questions with distant supervision, and even (3) do zero-shot learning by extracting new relation types that are only specified at test-time, for which we have no labeled training examples. Experiments on a Wikipedia slot-filling task demonstrate that the approach can generalize to new questions for known relation types with high accuracy, and that zero-shot generalization to unseen relation types is possible, at lower accuracy levels, setting the bar for future work on this task.

This paper is concerned with whether deep syntactic information can help surface parsing, with a particular focus on empty categories. We design new algorithms to produce dependency trees in which empty elements are allowed, and evaluate the impact of information about empty category on parsing overt elements. Such information is helpful to reduce the approximation error in a structured parsing model, but increases the search space for inference and accordingly the estimation error. To deal with structure-based overfitting, we propose to integrate disambiguation models with and without empty elements, and perform structure regularization via joint decoding. Experiments on English and Chinese TreeBanks with different parsing models indicate that incorporating empty elements consistently improves surface parsing.

This paper explores the information-theoretic measure entropy to detect metaphoric change, transferring ideas from hypernym detection to research on language change. We build the first diachronic test set for German as a standard for metaphoric change annotation. Our model is unsupervised, language-independent and generalizable to other processes of semantic change.

We study the representation and encoding of phonemes in a recurrent neural network model of grounded speech. We use a model which processes images and their spoken descriptions, and projects the visual and auditory representations into the same semantic space. We perform a number of analyses on how information about individual phonemes is encoded in the MFCC features extracted from the speech signal, and the activations of the layers of the model. Via experiments with phoneme decoding and phoneme discrimination we show that phoneme representations are most salient in the lower layers of the model, where low-level signals are processed at a fine-grained level, although a large amount of phonological information is retain at the top recurrent layer. We further find out that the attention mechanism following the top recurrent layer significantly attenuates encoding of phonology and makes the utterance embeddings much more invariant to synonymy. Moreover, a hierarchical clustering of phoneme representations learned by the network shows an organizational structure of phonemes similar to those proposed in linguistics.

Extending semantic parsing systems to new domains and languages is a highly expensive, time-consuming process, so making effective use of existing resources is critical. In this paper, we describe a transfer learning method using crosslingual word embeddings in a sequence-to-sequence model. On the NLmaps corpus, our approach achieves state-of-the-art accuracy of 85.7% for English. Most importantly, we observed a consistent improvement for German compared with several baseline domain adaptation techniques. As a by-product of this approach, our models that are trained on a combination of English and German utterances perform reasonably well on code-switching utterances which contain a mixture of English and German, even though the training data does not contain any such. As far as we know, this is the first study of code-switching in semantic parsing. We manually constructed the set of code-switching test utterances for the NLmaps corpus and achieve 78.3% accuracy on this dataset.

Coreference evaluation metrics are hard to optimize directly as they are non-differentiable functions, not easily decomposable into elementary decisions. Consequently, most approaches optimize objectives only indirectly related to the end goal, resulting in suboptimal performance. Instead, we propose a differentiable relaxation that lends itself to gradient-based optimisation, thus bypassing the need for reinforcement learning or heuristic modification of cross-entropy. We show that by modifying the training objective of a competitive neural coreference system, we obtain a substantial gain in performance. This suggests that our approach can be regarded as a viable alternative to using reinforcement learning or more computationally expensive imitation learning.

We introduce a neural network model that marries together ideas from two prominent strands of research on domain adaptation through representation learning: structural correspondence learning (SCL, (Blitzer et al., 2006)) and autoencoder neural networks (NNs). Our model is a three-layer NN that learns to encode the non-pivot features of an input example into a low dimensional representation, so that the existence of pivot features (features that are prominent in both domains and convey useful information for the NLP task) in the example can be decoded from that representation. The low-dimensional representation is then employed in a learning algorithm for the task. Moreover, we show how to inject pre-trained word embeddings into our model in order to improve generalization across examples with similar pivot features. We experiment with the task of cross-domain sentiment classification on 16 domain pairs and show substantial improvements over strong baselines.

We introduce a simple and accurate neural model for dependency-based semantic role labeling. Our model predicts predicate-argument dependencies relying on states of a bidirectional LSTM encoder. The semantic role labeler achieves competitive performance on English, even without any kind of syntactic information and only using local inference. However, when automatically predicted part-of-speech tags are provided as input, it substantially outperforms all previous local models and approaches the best reported results on the English CoNLL-2009 dataset. We also consider Chinese, Czech and Spanish where our approach also achieves competitive results. Syntactic parsers are unreliable on out-of-domain data, so standard (i.e., syntactically-informed) SRL models are hindered when tested in this setting. Our syntax-agnostic model appears more robust, resulting in the best reported results on standard out-of-domain test sets.

Part-of-speech (POS) tagging for morphologically rich languages such as Arabic is a challenging problem because of their enormous tag sets. One reason for this is that in the tagging scheme for such languages, a complete POS tag is formed by combining tags from multiple tag sets defined for each morphosyntactic category. Previous approaches in Arabic POS tagging applied one model for each morphosyntactic tagging task, without utilizing shared information between the tasks. In this paper, we propose an approach that utilizes this information by jointly modeling multiple morphosyntactic tagging tasks with a multi-task learning framework. We also propose a method of incorporating tag dictionary information into our neural models by combining word representations with representations of the sets of possible tags. Our experiments showed that the joint model with tag dictionary information results in an accuracy of 91.38% on the Penn Arabic Treebank data set, with an absolute improvement of 2.11% over the current state-of-the-art tagger.

Arabic dialects do not just share a common koiné, but there are shared pan-dialectal linguistic phenomena that allow computational models for dialects to learn from each other. In this paper we build a unified segmentation model where the training data for different dialects are combined and a single model is trained. The model yields higher accuracies than dialect-specific models, eliminating the need for dialect identification before segmentation. We also measure the degree of relatedness between four major Arabic dialects by testing how a segmentation model trained on one dialect performs on the other dialects. We found that linguistic relatedness is contingent with geographical proximity. In our experiments we use SVM-based ranking and bi-LSTM-CRF sequence labeling.

Natural language generation (NLG) is a critical component in a spoken dialogue system. This paper presents a Recurrent Neural Network based Encoder-Decoder architecture, in which an LSTM-based decoder is introduced to select, aggregate semantic elements produced by an attention mechanism over the input elements, and to produce the required utterances. The proposed generator can be jointly trained both sentence planning and surface realization to produce natural language sentences. The proposed model was extensively evaluated on four different NLG datasets. The experimental results showed that the proposed generators not only consistently outperform the previous methods across all the NLG domains but also show an ability to generalize from a new, unseen domain and learn from multi-domain datasets.

We propose a neural multi-document summarization system that incorporates sentence relation graphs. We employ a Graph Convolutional Network (GCN) on the relation graphs, with sentence embeddings obtained from Recurrent Neural Networks as input node features. Through multiple layer-wise propagation, the GCN generates high-level hidden sentence features for salience estimation. We then use a greedy heuristic to extract salient sentences that avoid redundancy. In our experiments on DUC 2004, we consider three types of sentence relation graphs and demonstrate the advantage of combining sentence relations in graphs with the representation power of deep neural networks. Our model improves upon other traditional graph-based extractive approaches and the vanilla GRU sequence model with no graph, and it achieves competitive results against other state-of-the-art multi-document summarization systems.