arXiv subjects

Mo Yu

Publications and source records attributed to Mo Yu.

At least 127 records · Page 7Linked to original sources

arXiv2018

Deriving Machine Attention from Human Rationales

Attention-based models are successful when trained on large amounts of data. In this paper, we demonstrate that even in the low-resource scenario, attention can be learned effectively. To this end, we start with discrete human-annotated rationales and map them into continuous attention. Our central hypothesis is that this mapping is general across domains, and thus can be transferred from resource-rich domains to low-resource ones. Our model jointly learns a domain-invariant representation and induces the desired mapping between rationales and attention. Our empirical results validate this hypothesis and show that our approach delivers significant gains over state-of-the-art baselines, yielding over 15% average error reduction on benchmark datasets.

cs.CL↗

arXiv2018

One-Shot Relational Learning for Knowledge Graphs

Knowledge graphs (KGs) are the key components of various natural language processing applications. To further expand KGs' coverage, previous studies on knowledge graph completion usually require a large number of training instances for each relation. However, we observe that long-tail relations are actually more common in KGs and those newly added relations often do not have many known triples for training. In this work, we aim at predicting new facts under a challenging setting where only one training instance is available. We propose a one-shot relational learning framework, which utilizes the knowledge extracted by embedding models and learns a matching metric by considering both the learned embeddings and one-hop graph structures. Empirically, our model yields considerable performance improvements over existing embedding models, and also eliminates the need of re-training the embedding models when dealing with newly added relations.

cs.CL↗

arXiv2018

Exploiting Rich Syntactic Information for Semantic Parsing with Graph-to-Sequence Model

Existing neural semantic parsers mainly utilize a sequence encoder, i.e., a sequential LSTM, to extract word order features while neglecting other valuable syntactic information such as dependency graph or constituent trees. In this paper, we first propose to use the \textit{syntactic graph} to represent three types of syntactic information, i.e., word order, dependency and constituency features. We further employ a graph-to-sequence model to encode the syntactic graph and decode a logical form. Experimental results on benchmark datasets show that our model is comparable to the state-of-the-art on Jobs640, ATIS and Geo880. Experimental results on adversarial examples demonstrate the robustness of the model is also improved by encoding more syntactic information.

cs.CL↗

arXiv2018

Spatial-Temporal Synergic Residual Learning for Video Person Re-Identification

We tackle the problem of person re-identification in video setting in this paper, which has been viewed as a crucial task in many applications. Meanwhile, it is very challenging since the task requires learning effective representations from video sequences with heterogeneous spatial-temporal information. We present a novel method - Spatial-Temporal Synergic Residual Network (STSRN) for this problem. STSRN contains a spatial residual extractor, a temporal residual processor and a spatial-temporal smooth module. The smoother can alleviate sample noises along the spatial-temporal dimensions thus enable STSRN extracts more robust spatial-temporal features of consecutive frames. Extensive experiments are conducted on several challenging datasets including iLIDS-VID, PRID2011 and MARS. The results demonstrate that the proposed method achieves consistently superior performance over most of state-of-the-art methods.

cs.CV↗

arXiv2018

Scheduled Policy Optimization for Natural Language Communication with Intelligent Agents

We investigate the task of learning to follow natural language instructions by jointly reasoning with visual observations and language inputs. In contrast to existing methods which start with learning from demonstrations (LfD) and then use reinforcement learning (RL) to fine-tune the model parameters, we propose a novel policy optimization algorithm which dynamically schedules demonstration learning and RL. The proposed training paradigm provides efficient exploration and better generalization beyond existing methods. Comparing to existing ensemble models, the best single model based on our proposed method tremendously decreases the execution error by over 50% on a block-world environment. To further illustrate the exploration strategy of our RL algorithm, We also include systematic studies on the evolution of policy entropy during training.

cs.CL↗

arXiv2018

A Co-Matching Model for Multi-choice Reading Comprehension

Multi-choice reading comprehension is a challenging task, which involves the matching between a passage and a question-answer pair. This paper proposes a new co-matching approach to this problem, which jointly models whether a passage can match both a question and a candidate answer. Experimental results on the RACE dataset demonstrate that our approach achieves state-of-the-art performance.

cs.CL↗

arXiv2018

Diverse Few-Shot Text Classification with Multiple Metrics

We study few-shot learning in natural language domains. Compared to many existing works that apply either metric-based or optimization-based meta-learning to image domain with low inter-task variance, we consider a more realistic setting, where tasks are diverse. However, it imposes tremendous difficulties to existing state-of-the-art metric-based algorithms since a single metric is insufficient to capture complex task variations in natural language domain. To alleviate the problem, we propose an adaptive metric learning approach that automatically determines the best weighted combination from a set of metrics obtained from meta-training tasks for a newly seen few-shot task. Extensive quantitative evaluations on real-world sentiment analysis and dialog intent classification datasets demonstrate that the proposed method performs favorably against state-of-the-art few shot learning algorithms in terms of predictive accuracy. We make our code and data available for further study.

cs.CL↗

arXiv2018

Robust Task Clustering for Deep Many-Task Learning

We investigate task clustering for deep-learning based multi-task and few-shot learning in a many-task setting. We propose a new method to measure task similarities with cross-task transfer performance matrix for the deep learning scenario. Although this matrix provides us critical information regarding similarity between tasks, its asymmetric property and unreliable performance scores can affect conventional clustering methods adversely. Additionally, the uncertain task-pairs, i.e., the ones with extremely asymmetric transfer scores, may collectively mislead clustering algorithms to output an inaccurate task-partition. To overcome these limitations, we propose a novel task-clustering algorithm by using the matrix completion technique. The proposed algorithm constructs a partially-observed similarity matrix based on the certainty of cluster membership of the task-pairs. We then use a matrix completion algorithm to complete the similarity matrix. Our theoretical analysis shows that under mild constraints, the proposed algorithm will perfectly recover the underlying "true" similarity matrix with a high probability. Our results show that the new task clustering method can discover task clusters for training flexible and superior neural network models in a multi-task learning setup for sentiment classification and dialog intent classification tasks. Our task clustering approach also extends metric-based few-shot learning methods to adapt multiple metrics, which demonstrates empirical advantages when the tasks are diverse.

cs.LG↗

arXiv2018

Image Super-Resolution via Dual-State Recurrent Networks

Advances in image super-resolution (SR) have recently benefited significantly from rapid developments in deep neural networks. Inspired by these recent discoveries, we note that many state-of-the-art deep SR architectures can be reformulated as a single-state recurrent neural network (RNN) with finite unfoldings. In this paper, we explore new structures for SR based on this compact RNN view, leading us to a dual-state design, the Dual-State Recurrent Network (DSRN). Compared to its single state counterparts that operate at a fixed spatial resolution, DSRN exploits both low-resolution (LR) and high-resolution (HR) signals jointly. Recurrent signals are exchanged between these states in both directions (both LR to HR and HR to LR) via delayed feedback. Extensive quantitative and qualitative evaluations on benchmark datasets and on a recent challenge demonstrate that the proposed DSRN performs favorably against state-of-the-art algorithms in terms of both memory consumption and predictive accuracy.

cs.CV↗

arXiv2018

Evidence Aggregation for Answer Re-Ranking in Open-Domain Question Answering

A popular recent approach to answering open-domain questions is to first search for question-related passages and then apply reading comprehension models to extract answers. Existing methods usually extract answers from single passages independently. But some questions require a combination of evidence from across different sources to answer correctly. In this paper, we propose two models which make use of multiple passages to generate their answers. Both use an answer-reranking approach which reorders the answer candidates generated by an existing state-of-the-art QA model. We propose two methods, namely, strength-based re-ranking and coverage-based re-ranking, to make use of the aggregated evidence from different passages to better determine the answer. Our models have achieved state-of-the-art results on three public open-domain QA datasets: Quasar-T, SearchQA and the open-domain version of TriviaQA, with about 8 percentage points of improvement over the former two datasets.

cs.CL↗

arXiv2018

Knowledge Base Relation Detection via Multi-View Matching

Relation detection is a core component for Knowledge Base Question Answering (KBQA). In this paper, we propose a KB relation detection model via multi-view matching which utilizes more useful information extracted from question and KB. The matching inside each view is through multiple perspectives to compare two input texts thoroughly. All these components are designed in an end-to-end trainable neural network model. Experiments on SimpleQuestions and WebQSP yield state-of-the-art results.

cs.AI↗

arXiv2017

R$^3$: Reinforced Reader-Ranker for Open-Domain Question Answering

In recent years researchers have achieved considerable success applying neural network methods to question answering (QA). These approaches have achieved state of the art results in simplified closed-domain settings such as the SQuAD (Rajpurkar et al., 2016) dataset, which provides a pre-selected passage, from which the answer to a given question may be extracted. More recently, researchers have begun to tackle open-domain QA, in which the model is given a question and access to a large corpus (e.g., wikipedia) instead of a pre-selected passage (Chen et al., 2017a). This setting is more complex as it requires large-scale search for relevant passages by an information retrieval component, combined with a reading comprehension model that "reads" the passages to generate an answer to the question. Performance in this setting lags considerably behind closed-domain performance. In this paper, we present a novel open-domain QA system called Reinforced Ranker-Reader $(R^3)$, based on two algorithmic innovations. First, we propose a new pipeline for open-domain QA with a Ranker component, which learns to rank retrieved passages in terms of likelihood of generating the ground-truth answer to a given question. Second, we propose a novel method that jointly trains the Ranker along with an answer-generation Reader model, based on reinforcement learning. We report extensive experimental results showing that our method significantly improves on the state of the art for multiple open-domain QA datasets.

cs.CL↗

arXiv2017

The Effects of Protostellar Disk Turbulence on CO Emission Lines: A Comparison Study of Disks with Constant CO Abundance vs. Chemically Evolving Disks

Turbulence is the leading candidate for angular momentum transport in protoplanetary disks and therefore influences disk lifetimes and planet formation timescales. However, the turbulent properties of protoplanetary disks are poorly constrained observationally. Recent studies have found turbulent speeds smaller than what fully-developed MRI would produce (Flaherty et al. 2015, 2017). However, existing studies assumed a constant CO/H2 ratio of 0.0001 in locations where CO is not frozen-out or photo-dissociated. Our previous studies of evolving disk chemistry indicate that CO is depleted by incorporation into complex organic molecules well inside the freeze-out radius of CO. We consider the effects of this chemical depletion on measurements of turbulence. Simon et al. (2015) suggested that the ratio of the peak line flux to the flux at line center of the CO J=3-2 transition is a reasonable diagnostic of turbulence, so we focus on that metric, while adding some analysis of the more complex effects on spatial distribution. We simulate the emission lines of CO based on chemical evolution models presented in Yu et al. (2016), and find that the peak-to-trough ratio changes as a function of time as CO is destroyed. Specifically, a CO-depleted disk with high turbulent velocity mimics the peak-to-trough ratios of a non-CO-depleted disk with lower turbulent velocity. We suggest that disk observers and modelers take into account the possibility of CO depletion when using line peak-to-trough ratios to constrain the degree of turbulence in disks. Assuming that CO/H2 = 0.0001 at all disk radii can lead to underestimates of turbulent speeds in the disk by at least 0.2 km/s.

astro-ph.EP↗

arXiv2017

Dilated Recurrent Neural Networks

Learning with recurrent neural networks (RNNs) on long sequences is a notoriously difficult task. There are three major challenges: 1) complex dependencies, 2) vanishing and exploding gradients, and 3) efficient parallelization. In this paper, we introduce a simple yet effective RNN connection structure, the DilatedRNN, which simultaneously tackles all of these challenges. The proposed architecture is characterized by multi-resolution dilated recurrent skip connections and can be combined flexibly with diverse RNN cells. Moreover, the DilatedRNN reduces the number of parameters needed and enhances training efficiency significantly, while matching state-of-the-art performance (even with standard RNN cells) in tasks involving very long-term dependencies. To provide a theory-based quantification of the architecture's advantages, we introduce a memory capacity measure, the mean recurrent length, which is more suitable for RNNs with long skip connections than existing measures. We rigorously prove the advantages of the DilatedRNN over other recurrent neural architectures. The code for our method is publicly available at https://github.com/code-terminator/DilatedRNN

cs.AI↗

arXiv2017

Improved Neural Relation Detection for Knowledge Base Question Answering

Relation detection is a core component for many NLP applications including Knowledge Base Question Answering (KBQA). In this paper, we propose a hierarchical recurrent neural network enhanced by residual learning that detects KB relations given an input question. Our method uses deep residual bidirectional LSTMs to compare questions and relation names via different hierarchies of abstraction. Additionally, we propose a simple KBQA system that integrates entity linking and our proposed relation detector to enable one enhance another. Experimental results evidence that our approach achieves not only outstanding relation detection performance, but more importantly, it helps our KBQA system to achieve state-of-the-art accuracy for both single-relation (SimpleQuestions) and multi-relation (WebQSP) QA benchmarks.

cs.CL↗

arXiv2017

Disk masses around solar-mass stars are underestimated by CO observations

Gas in protostellar disks provides the raw material for giant planet formation and controls the dynamics of the planetesimal-building dust grains. Accurate gas mass measurements help map the observed properties of planet-forming disks onto the formation environments of known exoplanets. Rare isotopologues of carbon monoxide (CO) have been used as gas mass tracers for disks in the Lupus star-forming region, with an assumed interstellar CO/H$_2$ abundance ratio. Unfortunately, observations of T-Tauri disks show that CO abundance is not interstellar---a finding reproduced by models that show CO abundance decreasing both with distance from the star and as a function of time. Here we present radiative transfer simulations that assess the accuracy of CO-based disk mass measurements. We find that the combination of CO chemical depletion in the outer disk and optically thick emission from the inner disk leads observers to underestimate gas mass by more than an order of magnitude if they use the standard assumptions of interstellar CO/H$_2$ ratio and optically thin emission. Furthermore, CO abundance changes on million-year timescales, introducing an age/mass degeneracy into observations. To reach factor of a few accuracy for CO-based disk mass measurements, we suggest that observers and modelers adopt the following strategies: (1) select the low-$J$ transitions; (2) observe multiple CO isotopologues and use either intensity ratios or normalized line profiles to diagnose CO chemical depletion; and (3) use spatially resolved observations to measure the CO abundance distribution.

astro-ph.EP↗

arXiv2017

A Structured Self-attentive Sentence Embedding

This paper proposes a new model for extracting an interpretable sentence embedding by introducing self-attention. Instead of using a vector, we use a 2-D matrix to represent the embedding, with each row of the matrix attending on a different part of the sentence. We also propose a self-attention mechanism and a special regularization term for the model. As a side effect, the embedding comes with an easy way of visualizing what specific parts of the sentence are encoded into the embedding. We evaluate our model on 3 different tasks: author profiling, sentiment classification, and textual entailment. Results show that our model yields a significant performance gain compared to other sentence embedding methods in all of the 3 tasks.

cs.CL↗

arXiv2017

Comparative Study of CNN and RNN for Natural Language Processing

Deep neural networks (DNN) have revolutionized the field of natural language processing (NLP). Convolutional neural network (CNN) and recurrent neural network (RNN), the two main types of DNN architectures, are widely explored to handle various NLP tasks. CNN is supposed to be good at extracting position-invariant features and RNN at modeling units in sequence. The state of the art on many NLP tasks often switches due to the battle between CNNs and RNNs. This work is the first systematic comparison of CNN and RNN on a wide range of representative NLP tasks, aiming to give basic guidance for DNN selection.

cs.CL↗