October 15, 2019

2124 words 10 mins read

Paper Group NANR 184

Paper Group NANR 184

Higher-Order Syntactic Attention Network for Longer Sentence Compression. A supervised approach to taxonomy extraction using word embeddings. Detecting Sarcasm is Extremely Easy ;-). Few-Shot Human Motion Prediction via Meta-Learning. A Survey of Machine Translation Work in the Philippines: From 1998 to 2018. MorAz: an Open-source Morphological Ana …

Higher-Order Syntactic Attention Network for Longer Sentence Compression

Title Higher-Order Syntactic Attention Network for Longer Sentence Compression
Authors Hidetaka Kamigaito, Katsuhiko Hayashi, Tsutomu Hirao, Masaaki Nagata
Abstract A sentence compression method using LSTM can generate fluent compressed sentences. However, the performance of this method is significantly degraded when compressing longer sentences since it does not explicitly handle syntactic features. To solve this problem, we propose a higher-order syntactic attention network (HiSAN) that can handle higher-order dependency features as an attention distribution on LSTM hidden states. Furthermore, to avoid the influence of incorrect parse results, we trained HiSAN by maximizing jointly the probability of a correct output with the attention distribution. Experimental results on Google sentence compression dataset showed that our method achieved the best performance on F1 as well as ROUGE-1,2 and L scores, 83.2, 82.9, 75.8 and 82.7, respectively. In human evaluation, our methods also outperformed baseline methods in both readability and informativeness.
Tasks Machine Translation, Sentence Compression
Published 2018-06-01
URL https://www.aclweb.org/anthology/N18-1155/
PDF https://www.aclweb.org/anthology/N18-1155
PWC https://paperswithcode.com/paper/higher-order-syntactic-attention-network-for
Repo
Framework

A supervised approach to taxonomy extraction using word embeddings

Title A supervised approach to taxonomy extraction using word embeddings
Authors Rajdeep Sarkar, John P. McCrae, Paul Buitelaar
Abstract
Tasks Word Embeddings
Published 2018-05-01
URL https://www.aclweb.org/anthology/L18-1324/
PDF https://www.aclweb.org/anthology/L18-1324
PWC https://paperswithcode.com/paper/a-supervised-approach-to-taxonomy-extraction
Repo
Framework

Detecting Sarcasm is Extremely Easy ;-)

Title Detecting Sarcasm is Extremely Easy ;-)
Authors Natalie Parde, Rodney Nielsen
Abstract Detecting sarcasm in text is a particularly challenging problem in computational semantics, and its solution may vary across different types of text. We analyze the performance of a domain-general sarcasm detection system on datasets from two very different domains: Twitter, and Amazon product reviews. We categorize the errors that we identify with each, and make recommendations for addressing these issues in NLP systems in the future.
Tasks Sarcasm Detection
Published 2018-06-01
URL https://www.aclweb.org/anthology/W18-1303/
PDF https://www.aclweb.org/anthology/W18-1303
PWC https://paperswithcode.com/paper/detecting-sarcasm-is-extremely-easy-
Repo
Framework

Few-Shot Human Motion Prediction via Meta-Learning

Title Few-Shot Human Motion Prediction via Meta-Learning
Authors Liang-Yan Gui, Yu-Xiong Wang, Deva Ramanan, Jose M. F. Moura
Abstract Human motion prediction, forecasting human motion in a few milliseconds conditioning on a historical 3D skeleton sequence, is a long-standing problem in computer vision and robotic vision. Existing forecasting algorithms rely on extensive annotated motion capture data and are brittle to novel actions. This paper addresses the problem of few-shot human motion prediction, in the spirit of the recent progress on few-shot learning and meta-learning. More precisely, our approach is based on the insight that having a good generalization from few examples relies on both a generic initial model and an effective strategy for adapting this model to novel tasks. To accomplish this, we propose proactive and adaptive meta-learning (PAML) that introduces a novel combination of model-agnostic meta-learning and model regression networks and unifies them into an integrated, end-to-end framework. By doing so, our meta-learner produces a generic model initialization through aggregating contextual information from a variety of prediction tasks, while this model can effectively adapt to a specific task by leveraging learning-to-learn knowledge about how to transform few-shot model parameters to many-shot model parameters. The resulting PAML predictor model significantly improves the prediction performance on the heavily benchmarked H3.6M dataset in the small-sample size regime.
Tasks Few-Shot Learning, Meta-Learning, Motion Capture, motion prediction
Published 2018-09-01
URL http://openaccess.thecvf.com/content_ECCV_2018/html/Liangyan_Gui_Few-Shot_Human_Motion_ECCV_2018_paper.html
PDF http://openaccess.thecvf.com/content_ECCV_2018/papers/Liangyan_Gui_Few-Shot_Human_Motion_ECCV_2018_paper.pdf
PWC https://paperswithcode.com/paper/few-shot-human-motion-prediction-via-meta
Repo
Framework

A Survey of Machine Translation Work in the Philippines: From 1998 to 2018

Title A Survey of Machine Translation Work in the Philippines: From 1998 to 2018
Authors Nathaniel Oco, Rachel Roxas
Abstract
Tasks Machine Translation
Published 2018-03-01
URL https://www.aclweb.org/anthology/W18-2204/
PDF https://www.aclweb.org/anthology/W18-2204
PWC https://paperswithcode.com/paper/a-survey-of-machine-translation-work-in-the
Repo
Framework

MorAz: an Open-source Morphological Analyzer for Azerbaijani Turkish

Title MorAz: an Open-source Morphological Analyzer for Azerbaijani Turkish
Authors Berke {"O}zen{\c{c}}, Razieh Ehsani, Ercan Solak
Abstract MorAz is an open-source morphological analyzer for Azerbaijani Turkish. The analyzer is available through both as a website for interactive exploration and as a RESTful web service for integration into a natural language processing pipeline. MorAz implements the morphology of Azerbaijani Turkish in two-level using Helsinki finite-state transducer and wraps the analyzer with python scripts in a Django instance.
Tasks Morphological Analysis
Published 2018-11-01
URL https://www.aclweb.org/anthology/D18-2005/
PDF https://www.aclweb.org/anthology/D18-2005
PWC https://paperswithcode.com/paper/moraz-an-open-source-morphological-analyzer
Repo
Framework

LEAP: Learning Embeddings for Adaptive Pace

Title LEAP: Learning Embeddings for Adaptive Pace
Authors Vithursan Thangarasa, Graham W. Taylor
Abstract Determining the optimal order in which data examples are presented to Deep Neural Networks during training is a non-trivial problem. However, choosing a non-trivial scheduling method may drastically improve convergence. In this paper, we propose a Self-Paced Learning (SPL)-fused Deep Metric Learning (DML) framework, which we call Learning Embeddings for Adaptive Pace (LEAP). Our method parameterizes mini-batches dynamically based on the \textit{easiness} and \textit{true diverseness} of the sample within a salient feature representation space. In LEAP, we train an \textit{embedding} Convolutional Neural Network (CNN) to learn an expressive representation space by adaptive density discrimination using the Magnet Loss. The \textit{student} CNN classifier dynamically selects samples to form a mini-batch based on the \textit{easiness} from cross-entropy losses and \textit{true diverseness} of examples from the representation space sculpted by the \textit{embedding} CNN. We evaluate LEAP using deep CNN architectures for the task of supervised image classification on MNIST, FashionMNIST, CIFAR-10, CIFAR-100, and SVHN. We show that the LEAP framework converges faster with respect to the number of mini-batch updates required to achieve a comparable or better test performance on each of the datasets.
Tasks Image Classification, Metric Learning
Published 2018-01-01
URL https://openreview.net/forum?id=rk9kKMZ0-
PDF https://openreview.net/pdf?id=rk9kKMZ0-
PWC https://paperswithcode.com/paper/leap-learning-embeddings-for-adaptive-pace
Repo
Framework

Multilingual Named Entity Recognition on Spanish-English Code-switched Tweets using Support Vector Machines

Title Multilingual Named Entity Recognition on Spanish-English Code-switched Tweets using Support Vector Machines
Authors Daniel Claeser, Samantha Kent, Dennis Felske
Abstract This paper describes our system submission for the ACL 2018 shared task on named entity recognition (NER) in code-switched Twitter data. Our best result (F1 = 53.65) was obtained using a Support Vector Machine (SVM) with 14 features combined with rule-based post processing.
Tasks Entity Linking, Named Entity Recognition
Published 2018-07-01
URL https://www.aclweb.org/anthology/W18-3218/
PDF https://www.aclweb.org/anthology/W18-3218
PWC https://paperswithcode.com/paper/multilingual-named-entity-recognition-on
Repo
Framework

A Leveled Reading Corpus of Modern Standard Arabic

Title A Leveled Reading Corpus of Modern Standard Arabic
Authors Muhamed Al Khalil, Hind Saddiki, Nizar Habash, Latifa Alfalasi
Abstract
Tasks Document Classification, Machine Translation, Part-Of-Speech Tagging, Sentiment Analysis, Speech Recognition, Text Simplification
Published 2018-05-01
URL https://www.aclweb.org/anthology/L18-1366/
PDF https://www.aclweb.org/anthology/L18-1366
PWC https://paperswithcode.com/paper/a-leveled-reading-corpus-of-modern-standard
Repo
Framework

Integrated accounts of behavioral and neuroimaging data using flexible recurrent neural network models

Title Integrated accounts of behavioral and neuroimaging data using flexible recurrent neural network models
Authors Amir Dezfouli, Richard Morris, Fabio T. Ramos, Peter Dayan, Bernard Balleine
Abstract Neuroscience studies of human decision-making abilities commonly involve subjects completing a decision-making task while BOLD signals are recorded using fMRI. Hypotheses are tested about which brain regions mediate the effect of past experience, such as rewards, on future actions. One standard approach to this is model-based fMRI data analysis, in which a model is fitted to the behavioral data, i.e., a subject’s choices, and then the neural data are parsed to find brain regions whose BOLD signals are related to the model’s internal signals. However, the internal mechanics of such purely behavioral models are not constrained by the neural data, and therefore might miss or mischaracterize aspects of the brain. To address this limitation, we introduce a new method using recurrent neural network models that are flexible enough to be jointly fitted to the behavioral and neural data. We trained a model so that its internal states were suitably related to neural activity during the task, while at the same time its output predicted the next action a subject would execute. We then used the fitted model to create a novel visualization of the relationship between the activity in brain regions at different times following a reward and the choices the subject subsequently made. Finally, we validated our method using a previously published dataset. We found that the model was able to recover the underlying neural substrates that were discovered by explicit model engineering in the previous work, and also derived new results regarding the temporal pattern of brain activity.
Tasks Decision Making
Published 2018-12-01
URL http://papers.nips.cc/paper/7677-integrated-accounts-of-behavioral-and-neuroimaging-data-using-flexible-recurrent-neural-network-models
PDF http://papers.nips.cc/paper/7677-integrated-accounts-of-behavioral-and-neuroimaging-data-using-flexible-recurrent-neural-network-models.pdf
PWC https://paperswithcode.com/paper/integrated-accounts-of-behavioral-and
Repo
Framework

Computational Argumentation: A Journey Beyond Semantics, Logic, Opinions, and Easy Tasks

Title Computational Argumentation: A Journey Beyond Semantics, Logic, Opinions, and Easy Tasks
Authors Ivan Habernal
Abstract The classical view on argumentation, such that arguments are logical structures consisting of different distinguishable parts and that parties exchange arguments in a rational way, is prevalent in textbooks but nonexistent in the real world. Instead, argumentation is a multifaceted communication tool built upon humans{'} capabilities to easily use common sense, emotions, and social context. As humans, we are pretty good at it. Computational Argumentation tries to tackle these phenomena but has a long and not so easy way to go. In this talk, I would like to shed a light on several recent attempts to deal with argumentation computationally, such as addressing argument quality, understanding argument reasoning, dealing with fallacies, and how should we never ever argue online.
Tasks Common Sense Reasoning
Published 2018-06-01
URL https://www.aclweb.org/anthology/W18-1305/
PDF https://www.aclweb.org/anthology/W18-1305
PWC https://paperswithcode.com/paper/computational-argumentation-a-journey-beyond
Repo
Framework

Parametrized Hierarchical Procedures for Neural Programming

Title Parametrized Hierarchical Procedures for Neural Programming
Authors Roy Fox, Richard Shin, Sanjay Krishnan, Ken Goldberg, Dawn Song, Ion Stoica
Abstract Neural programs are highly accurate and structured policies that perform algorithmic tasks by controlling the behavior of a computation mechanism. Despite the potential to increase the interpretability and the compositionality of the behavior of artificial agents, it remains difficult to learn from demonstrations neural networks that represent computer programs. The main challenges that set algorithmic domains apart from other imitation learning domains are the need for high accuracy, the involvement of specific structures of data, and the extremely limited observability. To address these challenges, we propose to model programs as Parametrized Hierarchical Procedures (PHPs). A PHP is a sequence of conditional operations, using a program counter along with the observation to select between taking an elementary action, invoking another PHP as a sub-procedure, and returning to the caller. We develop an algorithm for training PHPs from a set of supervisor demonstrations, only some of which are annotated with the internal call structure, and apply it to efficient level-wise training of multi-level PHPs. We show in two benchmarks, NanoCraft and long-hand addition, that PHPs can learn neural programs more accurately from smaller amounts of both annotated and unannotated demonstrations.
Tasks Imitation Learning
Published 2018-01-01
URL https://openreview.net/forum?id=rJl63fZRb
PDF https://openreview.net/pdf?id=rJl63fZRb
PWC https://paperswithcode.com/paper/parametrized-hierarchical-procedures-for
Repo
Framework

FonBund: A Library for Combining Cross-lingual Phonological Segment Data

Title FonBund: A Library for Combining Cross-lingual Phonological Segment Data
Authors Alex Gutkin, er, Martin Jansche, Tatiana Merkulova
Abstract
Tasks Language Modelling, Speech Synthesis
Published 2018-05-01
URL https://www.aclweb.org/anthology/L18-1353/
PDF https://www.aclweb.org/anthology/L18-1353
PWC https://paperswithcode.com/paper/fonbund-a-library-for-combining-cross-lingual
Repo
Framework

Bayesian Model Selection Approach to Boundary Detection with Non-Local Priors

Title Bayesian Model Selection Approach to Boundary Detection with Non-Local Priors
Authors Fei Jiang, Guosheng Yin, Francesca Dominici
Abstract Based on non-local prior distributions, we propose a Bayesian model selection (BMS) procedure for boundary detection in a sequence of data with multiple systematic mean changes. The BMS method can effectively suppress the non-boundary spike points with large instantaneous changes. We speed up the algorithm by reducing the multiple change points to a series of single change point detection problems. We establish the consistency of the estimated number and locations of the change points under various prior distributions. Extensive simulation studies are conducted to compare the BMS with existing methods, and our approach is illustrated with application to the magnetic resonance imaging guided radiation therapy data.
Tasks Boundary Detection, Change Point Detection, Model Selection
Published 2018-12-01
URL http://papers.nips.cc/paper/7468-bayesian-model-selection-approach-to-boundary-detection-with-non-local-priors
PDF http://papers.nips.cc/paper/7468-bayesian-model-selection-approach-to-boundary-detection-with-non-local-priors.pdf
PWC https://paperswithcode.com/paper/bayesian-model-selection-approach-to-boundary
Repo
Framework

Recovering Missing Characters in Old Hawaiian Writing

Title Recovering Missing Characters in Old Hawaiian Writing
Authors Brendan Shillingford, Oiwi Parker Jones
Abstract In contrast to the older writing system of the 19th century, modern Hawaiian orthography employs characters for long vowels and glottal stops. These extra characters account for about one-third of the phonemes in Hawaiian, so including them makes a big difference to reading comprehension and pronunciation. However, transliterating between older and newer texts is a laborious task when performed manually. We introduce two related methods to help solve this transliteration problem automatically. One approach is implemented, end-to-end, using finite state transducers (FSTs). The other is a hybrid deep learning approach, which approximately composes an FST with a recurrent neural network language model.
Tasks Language Modelling, Reading Comprehension, Transliteration
Published 2018-10-01
URL https://www.aclweb.org/anthology/D18-1533/
PDF https://www.aclweb.org/anthology/D18-1533
PWC https://paperswithcode.com/paper/recovering-missing-characters-in-old-hawaiian
Repo
Framework
comments powered by Disqus