Paper Group NANR 205
Constructing a Lexicon of Relational Nouns. Textual Features Indicative of Writing Proficiency in Elementary School Spanish Documents. Link Weight Prediction with Node Embeddings. Tibetan-Chinese Neural Machine Translation based on Syllable Segmentation. Punctuation as Native Language Interference. Twitter Universal Dependency Parsing for African-A …
Constructing a Lexicon of Relational Nouns
| Title | Constructing a Lexicon of Relational Nouns |
| Authors | Edward Newell, Jackie C.K. Cheung |
| Abstract | |
| Tasks | Natural Language Inference, Question Answering, Relation Extraction |
| Published | 2018-05-01 |
| URL | https://www.aclweb.org/anthology/L18-1537/ |
| https://www.aclweb.org/anthology/L18-1537 | |
| PWC | https://paperswithcode.com/paper/constructing-a-lexicon-of-relational-nouns |
| Repo | |
| Framework | |
Textual Features Indicative of Writing Proficiency in Elementary School Spanish Documents
| Title | Textual Features Indicative of Writing Proficiency in Elementary School Spanish Documents |
| Authors | Gemma Bel-Enguix, Diana Due{~n}as Ch{'a}vez, Arturo Curiel D{'\i}az |
| Abstract | Childhood acquisition of written language is not straightforward. Writing skills evolve differently depending on external factors, such as the conditions in which children practice their productions and the quality of their instructors{'} guidance. This can be challenging in low-income areas, where schools may struggle to ensure ideal acquisition conditions. Developing computational tools to support the learning process may counterweight negative environmental influences; however, few work exists on the use of information technologies to improve childhood literacy. This work centers around the computational study of Spanish word and syllable structure in documents written by 2nd and 3rd year elementary school students. The studied texts were compared against a corpus of short stories aimed at the same age group, so as to observe whether the children tend to produce similar written patterns as the ones they are expected to interpret at their literacy level. The obtained results show some significant differences between the two kinds of texts, pointing towards possible strategies for the implementation of new education software in support of written language acquisition. |
| Tasks | Language Acquisition |
| Published | 2018-07-01 |
| URL | https://www.aclweb.org/anthology/W18-3716/ |
| https://www.aclweb.org/anthology/W18-3716 | |
| PWC | https://paperswithcode.com/paper/textual-features-indicative-of-writing |
| Repo | |
| Framework | |
Link Weight Prediction with Node Embeddings
| Title | Link Weight Prediction with Node Embeddings |
| Authors | Yuchen Hou, Lawrence B. Holder |
| Abstract | Application of deep learning has been successful in various domains such as im- age recognition, speech recognition and natural language processing. However, the research on its application in graph mining is still in an early stage. Here we present the first generic deep learning approach to the graph link weight prediction problem based on node embeddings. We evaluate this approach with three differ- ent node embedding techniques experimentally and compare its performance with two state-of-the-art non deep learning baseline approaches. Our experiment re- sults suggest that this deep learning approach outperforms the baselines by up to 70% depending on the dataset and embedding technique applied. This approach shows that deep learning can be successfully applied to link weight prediction to improve prediction accuracy. |
| Tasks | Speech Recognition |
| Published | 2018-01-01 |
| URL | https://openreview.net/forum?id=ryZ3KCy0W |
| https://openreview.net/pdf?id=ryZ3KCy0W | |
| PWC | https://paperswithcode.com/paper/link-weight-prediction-with-node-embeddings |
| Repo | |
| Framework | |
Tibetan-Chinese Neural Machine Translation based on Syllable Segmentation
| Title | Tibetan-Chinese Neural Machine Translation based on Syllable Segmentation |
| Authors | Wen Lai, Xiaobing Zhao, Wei Bao |
| Abstract | |
| Tasks | Machine Translation |
| Published | 2018-03-01 |
| URL | https://www.aclweb.org/anthology/W18-2203/ |
| https://www.aclweb.org/anthology/W18-2203 | |
| PWC | https://paperswithcode.com/paper/tibetan-chinese-neural-machine-translation |
| Repo | |
| Framework | |
Punctuation as Native Language Interference
| Title | Punctuation as Native Language Interference |
| Authors | Ilia Markov, Vivi Nastase, Carlo Strapparava |
| Abstract | In this paper, we describe experiments designed to explore and evaluate the impact of punctuation marks on the task of native language identification. Punctuation is specific to each language, and is part of the indicators that overtly represent the manner in which each language organizes and conveys information. Our experiments are organized in various set-ups: the usual multi-class classification for individual languages, also considering classification by language groups, across different proficiency levels, topics and even cross-corpus. The results support our hypothesis that punctuation marks are persistent and robust indicators of the native language of the author, which do not diminish in influence even when a high proficiency level in a non-native language is achieved. |
| Tasks | Language Identification, Native Language Identification |
| Published | 2018-08-01 |
| URL | https://www.aclweb.org/anthology/C18-1293/ |
| https://www.aclweb.org/anthology/C18-1293 | |
| PWC | https://paperswithcode.com/paper/punctuation-as-native-language-interference |
| Repo | |
| Framework | |
Twitter Universal Dependency Parsing for African-American and Mainstream American English
| Title | Twitter Universal Dependency Parsing for African-American and Mainstream American English |
| Authors | Su Lin Blodgett, Johnny Wei, Brendan O{'}Connor |
| Abstract | Due to the presence of both Twitter-specific conventions and non-standard and dialectal language, Twitter presents a significant parsing challenge to current dependency parsing tools. We broaden English dependency parsing to handle social media English, particularly social media African-American English (AAE), by developing and annotating a new dataset of 500 tweets, 250 of which are in AAE, within the Universal Dependencies 2.0 framework. We describe our standards for handling Twitter- and AAE-specific features and evaluate a variety of cross-domain strategies for improving parsing with no, or very little, in-domain labeled data, including a new data synthesis approach. We analyze these methods{'} impact on performance disparities between AAE and Mainstream American English tweets, and assess parsing accuracy for specific AAE lexical and syntactic features. Our annotated data and a parsing model are available at: \url{http://slanglab.cs.umass.edu/TwitterAAE/}. |
| Tasks | Dependency Parsing, Information Retrieval, Language Identification, Part-Of-Speech Tagging, Sentiment Analysis |
| Published | 2018-07-01 |
| URL | https://www.aclweb.org/anthology/P18-1131/ |
| https://www.aclweb.org/anthology/P18-1131 | |
| PWC | https://paperswithcode.com/paper/twitter-universal-dependency-parsing-for |
| Repo | |
| Framework | |
Optimal Tuning for Divide-and-conquer Kernel Ridge Regression with Massive Data
| Title | Optimal Tuning for Divide-and-conquer Kernel Ridge Regression with Massive Data |
| Authors | Ganggang Xu, Zuofeng Shang, Guang Cheng |
| Abstract | Divide-and-conquer is a powerful approach for large and massive data analysis. In the nonparameteric regression setting, although various theoretical frameworks have been established to achieve optimality in estimation or hypothesis testing, how to choose the tuning parameter in a practically effective way is still an open problem. In this paper, we propose a data-driven procedure based on divide-and-conquer for selecting the tuning parameters in kernel ridge regression by modifying the popular Generalized Cross-validation (GCV, Wahba, 1990). While the proposed criterion is computationally scalable for massive data sets, it is also shown under mild conditions to be asymptotically optimal in the sense that minimizing the proposed distributed-GCV (dGCV) criterion is equivalent to minimizing the true global conditional empirical loss of the averaged function estimator, extending the existing optimality results of GCV to the divide-and-conquer framework. |
| Tasks | |
| Published | 2018-07-01 |
| URL | https://icml.cc/Conferences/2018/Schedule?showEvent=2205 |
| http://proceedings.mlr.press/v80/xu18f/xu18f.pdf | |
| PWC | https://paperswithcode.com/paper/optimal-tuning-for-divide-and-conquer-kernel |
| Repo | |
| Framework | |
Multiscale Hidden Markov Models For Covariance Prediction
| Title | Multiscale Hidden Markov Models For Covariance Prediction |
| Authors | João Sedoc, Jordan Rodu, Dean Foster, Lyle Ungar |
| Abstract | This paper presents a novel variant of hierarchical hidden Markov models (HMMs), the multiscale hidden Markov model (MSHMM), and an associated spectral estimation and prediction scheme that is consistent, finds global optima, and is computationally efficient. Our MSHMM is a generative model of multiple HMMs evolving at different rates where the observation is a result of the additive emissions of the HMMs. While estimation is relatively straightforward, prediction for the MSHMM poses a unique challenge, which we address in this paper. Further, we show that spectral estimation of the MSHMM outperforms standard methods of predicting the asset covariance of stock prices, a widely addressed problem that is multiscale, non-stationary, and requires processing huge amounts of data. |
| Tasks | |
| Published | 2018-01-01 |
| URL | https://openreview.net/forum?id=ByW5yxgA- |
| https://openreview.net/pdf?id=ByW5yxgA- | |
| PWC | https://paperswithcode.com/paper/multiscale-hidden-markov-models-for |
| Repo | |
| Framework | |
NLPZZX at SemEval-2018 Task 1: Using Ensemble Method for Emotion and Sentiment Intensity Determination
| Title | NLPZZX at SemEval-2018 Task 1: Using Ensemble Method for Emotion and Sentiment Intensity Determination |
| Authors | Zhengxin Zhang, Qimin Zhou, Hao Wu |
| Abstract | In this paper, we put forward a system that competed at SemEval-2018 Task 1: {``}Affect in Tweets{''}. Our system uses a simple yet effective ensemble method which combines several neural network components. We participate in two subtasks for English tweets: EI-reg and V-reg. For two subtasks, different combinations of neural components are examined. For EI-reg, our system achieves an accuracy of 0.727 in Pearson Correlation Coefficient (all instances) and an accuracy of 0.555 in Pearson Correlation Coefficient (0.5-1). For V-reg, the achieved accuracy scores are respectively 0.835 and 0.670 | |
| Tasks | Sentiment Analysis |
| Published | 2018-06-01 |
| URL | https://www.aclweb.org/anthology/S18-1015/ |
| https://www.aclweb.org/anthology/S18-1015 | |
| PWC | https://paperswithcode.com/paper/nlpzzx-at-semeval-2018-task-1-using-ensemble |
| Repo | |
| Framework | |
Sparse-Complementary Convolution for Efficient Model Utilization on CNNs
| Title | Sparse-Complementary Convolution for Efficient Model Utilization on CNNs |
| Authors | Chun-Fu (Richard) Chen, Jinwook Oh, Quanfu Fan, Marco Pistoia, Gwo Giun (Chris) Lee |
| Abstract | We introduce an efficient way to increase the accuracy of convolution neural networks (CNNs) based on high model utilization without increasing any computational complexity. The proposed sparse-complementary convolution replaces regular convolution with sparse and complementary shapes of kernels, covering the same receptive field. By the nature of deep learning, high model utilization of a CNN can be achieved with more simpler kernels rather than fewer complex kernels. This simple but insightful model reuses of recent network architectures, ResNet and DenseNet, can provide better accuracy for most classification tasks (CIFAR-10/100 and ImageNet) compared to their baseline models. By simply replacing the convolution of a CNN with our sparse-complementary convolution, at the same FLOPs and parameters, we can improve top-1 accuracy on ImageNet by 0.33% and 0.18% for ResNet-101 and ResNet-152, respectively. A similar accuracy improvement could be gained by increasing the number of layers in those networks by ~1.5x. |
| Tasks | |
| Published | 2018-01-01 |
| URL | https://openreview.net/forum?id=HkeJVllRW |
| https://openreview.net/pdf?id=HkeJVllRW | |
| PWC | https://paperswithcode.com/paper/sparse-complementary-convolution-for |
| Repo | |
| Framework | |
Geometry Aware Constrained Optimization Techniques for Deep Learning
| Title | Geometry Aware Constrained Optimization Techniques for Deep Learning |
| Authors | Soumava Kumar Roy, Zakaria Mhammedi, Mehrtash Harandi |
| Abstract | In this paper, we generalize the Stochastic Gradient Descent (SGD) and RMSProp algorithms to the setting of Riemannian optimization. SGD is a popular method for large scale optimization. In particular, it is widely used to train the weights of Deep Neural Networks. However, gradients computed using standard SGD can have large variance, which is detrimental for the convergence rate of the algorithm. Other methods such as RMSProp and ADAM address this issue. Nevertheless, these methods cannot be directly applied to constrained optimization problems. In this paper, we extend some popular optimization algorithm to the Riemannian (constrained) setting. We substantiate our proposed extensions with a range of relevant problems in machine learning such as incremental Principal Component Analysis, computating the Riemannian centroids of SPD matrices, and Deep Metric Learning. We achieve competitive results against the state of the art for fine-grained object recognition datasets. |
| Tasks | Metric Learning, Object Recognition |
| Published | 2018-06-01 |
| URL | http://openaccess.thecvf.com/content_cvpr_2018/html/Roy_Geometry_Aware_Constrained_CVPR_2018_paper.html |
| http://openaccess.thecvf.com/content_cvpr_2018/papers/Roy_Geometry_Aware_Constrained_CVPR_2018_paper.pdf | |
| PWC | https://paperswithcode.com/paper/geometry-aware-constrained-optimization |
| Repo | |
| Framework | |
Learning Generative Models with Locally Disentangled Latent Factors
| Title | Learning Generative Models with Locally Disentangled Latent Factors |
| Authors | Brady Neal, Alex Lamb, Sherjil Ozair, Devon Hjelm, Aaron Courville, Yoshua Bengio, Ioannis Mitliagkas |
| Abstract | One of the most successful techniques in generative models has been decomposing a complicated generation task into a series of simpler generation tasks. For example, generating an image at a low resolution and then learning to refine that into a high resolution image often improves results substantially. Here we explore a novel strategy for decomposing generation for complicated objects in which we first generate latent variables which describe a subset of the observed variables, and then map from these latent variables to the observed space. We show that this allows us to achieve decoupled training of complicated generative models and present both theoretical and experimental results supporting the benefit of such an approach. |
| Tasks | |
| Published | 2018-01-01 |
| URL | https://openreview.net/forum?id=rJTGkKxAZ |
| https://openreview.net/pdf?id=rJTGkKxAZ | |
| PWC | https://paperswithcode.com/paper/learning-generative-models-with-locally |
| Repo | |
| Framework | |
Dependency Parsing of Code-Switching Data with Cross-Lingual Feature Representations
| Title | Dependency Parsing of Code-Switching Data with Cross-Lingual Feature Representations |
| Authors | Niko Partanen, Kyungtae Lim, Michael Rie{\ss}ler, Thierry Poibeau |
| Abstract | |
| Tasks | Dependency Parsing |
| Published | 2018-01-01 |
| URL | https://www.aclweb.org/anthology/W18-0201/ |
| https://www.aclweb.org/anthology/W18-0201 | |
| PWC | https://paperswithcode.com/paper/dependency-parsing-of-code-switching-data |
| Repo | |
| Framework | |
Adaptive Learning of Local Semantic and Global Structure Representations for Text Classification
| Title | Adaptive Learning of Local Semantic and Global Structure Representations for Text Classification |
| Authors | Jianyu Zhao, Zhiqiang Zhan, Qichuan Yang, Yang Zhang, Changjian Hu, Zhensheng Li, Liuxin Zhang, Zhiqiang He |
| Abstract | Representation learning is a key issue for most Natural Language Processing (NLP) tasks. Most existing representation models either learn little structure information or just rely on pre-defined structures, leading to degradation of performance and generalization capability. This paper focuses on learning both local semantic and global structure representations for text classification. In detail, we propose a novel Sandwich Neural Network (SNN) to learn semantic and structure representations automatically without relying on parsers. More importantly, semantic and structure information contribute unequally to the text representation at corpus and instance level. To solve the fusion problem, we propose two strategies: Adaptive Learning Sandwich Neural Network (AL-SNN) and Self-Attention Sandwich Neural Network (SA-SNN). The former learns the weights at corpus level, and the latter further combines attention mechanism to assign the weights at instance level. Experimental results demonstrate that our approach achieves competitive performance on several text classification tasks, including sentiment analysis, question type classification and subjectivity classification. Specifically, the accuracies are MR (82.1{%}), SST-5 (50.4{%}), TREC (96{%}) and SUBJ (93.9{%}). |
| Tasks | Representation Learning, Sentiment Analysis, Text Classification |
| Published | 2018-08-01 |
| URL | https://www.aclweb.org/anthology/C18-1173/ |
| https://www.aclweb.org/anthology/C18-1173 | |
| PWC | https://paperswithcode.com/paper/adaptive-learning-of-local-semantic-and |
| Repo | |
| Framework | |
DebugSL: An Interactive Tool for Debugging Sentiment Lexicons
| Title | DebugSL: An Interactive Tool for Debugging Sentiment Lexicons |
| Authors | Andrew Schneider, John Male, Saroja Bhogadhi, Eduard Dragut |
| Abstract | We introduce DebugSL, a visual (Web) debugging tool for sentiment lexicons (SLs). Its core component implements our algorithms for the automatic detection of polarity inconsistencies in SLs. An inconsistency is a set of words and/or word-senses whose polarity assignments cannot all be simultaneously satisfied. DebugSL finds inconsistencies of small sizes in SLs and has a rich user interface which helps users in the correction process. The project source code is available at \url{https://github.com/atschneid/DebugSL} A screencast of DebugSL can be viewed at \url{https://cis.temple.edu/~edragut/DebugSL.webm} |
| Tasks | Sentiment Analysis |
| Published | 2018-06-01 |
| URL | https://www.aclweb.org/anthology/N18-5008/ |
| https://www.aclweb.org/anthology/N18-5008 | |
| PWC | https://paperswithcode.com/paper/debugsl-an-interactive-tool-for-debugging |
| Repo | |
| Framework | |
