Read more from aclanthology.org

A Templatebased Abstractive Meeting Summarization: Leveraging Summary and Source Text Relationships · Tatsuro Oya Yashar Mehdad Giuseppe Carenini  - Information from aclanthology.org

Discourse processing is a suite of Natural Language Processing (NLP) tasks to uncover linguistic structures from texts at several levels which can support many  - Information from aclanthology.org

tionally processing the discontinuous constituents of non configurational languages In these languages the grammatical func - Information from aclanthology.org

Based on this idea we propose our Curriculum Learning approach By reviewing the trainset in a crossed way we are able to distinguish easy examples from  - Information from aclanthology.org

We investigate a simple yet effective method lexical substitution with word representation models to automatically expand a small set of frameannotated  - Information from aclanthology.org

hort identification phenotypespecific algorithms Active learning is an approach to selecting unla Bioinformatics 26(9):1205–1210 - Information from aclanthology.org

In this paper we describe a transformer model pretrained on 8 billion tokens of crawled text from the Croatian Bosnian Serbian and - Information from aclanthology.org

A Convolutional Neural Network for Modelling Sentences · Nal Kalchbrenner Edward Grefenstette Phil Blunsom Anthology ID: P141062 Volume: Proceedings of the  - Information from aclanthology.org

Relation Classification via MultiLevel Attention CNNs · Linlin Wang Zhu Cao Gerard de Melo Zhiyuan Liu Anthology ID: P161123 Volume: Proceedings of  - Information from aclanthology.org

(10) Prosodic Foot in Japanese (Moraic Trochees) Thus it claims that feet are built on syllables in all stress languages and moras as weight units are - Information from aclanthology.org

Figure 1: Multimodal Compact Bilinear Pooling for visual question answering 2015) and convolutional neural networks (CNNs) have shown to work best to  - Information from aclanthology.org

parallel RL training using asynchronous methods and propose RLIEA3C RLIE A3C trains multiple agents in parallel and is able to achieve upto 6x training - Information from aclanthology.org

nextword prediction in language modeling tasks Such subwordlevel informed models should be particularly effective for morphologicallyrich languages  - Information from aclanthology.org

Pavlo Kapustin 2020 pdf bib abs Semantic parsing with fuzzy meaning representations · Pavlo Kapustin Michael Kapustin - Information from aclanthology.org

Abstract We have adapted a classification approach coming from optical character recognition research to the task of speech emotion - Information from aclanthology.org

al sequence generative adversarial net which propose the Transformer the first sequence trans is a weighted combination of the naive GAN and MRT - Information from aclanthology.org

This paper investigates how to effectively incorporate a pretrained masked language model (MLM) such as BERT into an encoderdecoder (EncDec) model for  - Information from aclanthology.org

lenge of building a computer system that can per Figure 1: Overview of the DeepQA architecture Like the Watson system the project framework - Information from aclanthology.org

As a simpler solution we instead combine representations obtained from BERT and from the static word vectors at sentence level In particular to obtain a  - Information from aclanthology.org

In this paper we propose an additionsubtraction twingated recurrent network (ATR) to simplify neural machine translation The recurrent units of ATR are  - Information from aclanthology.org

a generator using either a recurrent neural network (RNN) (Yu et al 2017 ment our evaluation on several RNNbased GANs: rnnganeval - Information from aclanthology.org

Moreover it reduces adversarial example generation time by 19x39x We also show that our approach transfers to a blackbox setting by attacking The  - Information from aclanthology.org

see their README at githubcom/williamleif/ socialsent Department of Computer Science 6 Available online at https://cdliuclaedu/ - Information from aclanthology.org

This paper presents a worked exercise of paper reproduction with the goal of implementing the results of the recent Transformer model - Information from aclanthology.org

Graph convolutional networks (GCNs) have been applied recently to text classification and produced an excellent performance However existing GCNbased  - Information from aclanthology.org

Although there have been several other applications of this model for natu ral language representation (Arora et al 2017 Mu and Viswanath 2018) ours is  - Information from aclanthology.org

Expanding English and Chinese Dictionaries by Wikipedia Titles Often they share moti tic relaxation gibbs distributions and the bayesian - Information from aclanthology.org

Current multilingual word translation methods are focused on jointly learning mappings from each language to a shared space The actual translation however  - Information from aclanthology.org

Generative Adversarial Networks (GANs) are a generator using either a recurrent neural network GANbased text generation and existing metrics - Information from aclanthology.org

Build it Break it Fix it for Dialogue Safety: Robustness from Adversarial Human Attack · Emily Dinan Samuel Humeau Bharath Chintagunta Jason Weston  - Information from aclanthology.org

“Abdominal/pelvic lymph nodes: There is intraperi toneal and retroperitoneal lymphadenopathy for ex ample enlarged mesenteric/peripancreatic lymph - Information from aclanthology.org

Proceedings of the 12th Conference on Language Resources and Evaluation (LREC 2020) ARIADNE project as briefly described in the introduction - Information from aclanthology.org

Anthology ID: K152011 Volume: Proceedings of the Nineteenth Conference on Computational Natural Language Learning Shared Task Month: July Year: 2015  - Information from aclanthology.org

Figure 1: An example of linguistic factor annota tions for a source sentence in English tures to integrate these additional factors into the NMT framework - Information from aclanthology.org

In this paper we study yes/no questions that are naturally occurring — meaning that they are generated in unprompted and unconstrained - Information from aclanthology.org

2020 pdf bib abs The impact of preprint servers in the formation of novel ideas · Swarup Satish Zonghai Yao Andrew Drozdov Boris Veytsman - Information from aclanthology.org

It yields a principled Bayesian learning algorithm adding gradient noise during training (enhancing exploration of the ganetal2017scalable - Information from aclanthology.org

Visual Bilingual Lexicon Induction with Transferred ConvNet Features · Douwe Kiela Ivan Vulić Stephen Clark Anthology ID: D151015 Volume: Proceedings  - Information from aclanthology.org

processes the context document Prior work has investigated both unsupervised and supervised ap proaches to train the retriever - Information from aclanthology.org

WikiSQL is a newly released dataset for studying the natural language sequence to SQL translation problem The SQL queries in WikiSQL are simple: Each  - Information from aclanthology.org

WikiMatrix : Mining 135M Parallel Sentences in 1620 Language Pairs from Wikipedia Holger Schwenk Facebook AI schwenkfbcom - Information from aclanthology.org

bhardwajetal2010anveshan Copy Citation: BibTeX MODS XML Endnote More options… PDF: https://aclanthologyorg/W101806pdf - Information from aclanthology.org

We study the impact of using rich and diverse textual descriptions of classes for zeroshot learning (ZSL) on ImageNet We create a new dataset ImageNetWiki  - Information from aclanthology.org

We introduce a new language representation model called BERT which stands for Bidirectional Encoder Representations from Transformers - Information from aclanthology.org

~resupposition [I] and eonven¢ional implicature [6] The models used in traditional modeltheoretic semantics provide a complete and static representation  - Information from aclanthology.org

We investigate an alternative simple method to use monolingual data for NMT training: We combine the scores of a pretrained and fixed language model (LM)  - Information from aclanthology.org

Shun Ishizaki 2012 pdf bib · Analysis of Travel Review Data from Readers Point of View · Maya Ando Shun Ishizaki - Information from aclanthology.org

Convolutional Neural Networks for Sentence Classification Language: URL: https://aclanthologyorg/D141181 DOI: 103115/v1/D141181 Bibkey: - Information from aclanthology.org

Prior work treated animacy as a wordlevel property and has developed statistical classifiers to classify words as either animate or inanimate We discuss why  - Information from aclanthology.org

Building on the Word Movers Distance metric and var ious word embeddings we introduce a frag mentation penalty to account for fluency of - Information from aclanthology.org

Opinion Mining with Deep Recurrent Neural Networks · Ozan İrsoy Claire Cardie Anthology ID: D141080 Volume: Proceedings of the 2014 Conference on  - Information from aclanthology.org

Figure 1: Examples of images and image descrip tions from the Conceptual Captions dataset we start from existing alttext descriptions and auto matically  - Information from aclanthology.org

Figure 1: Greybox adversarial attack and defence framework for sentiment classification The training phase is divided into attacking steps and defending steps  - Information from aclanthology.org

Toward Future Scenario Generation: Extracting Event Causality Exploiting Semantic Relation Context and Association Features Chikara Hashimoto - Information from aclanthology.org

Graphics Processing Units (GPUs) are commonly used to train and evaluate neural networks efficiently While previous work in deep learning has focused on  - Information from aclanthology.org

Further we publicly release our Poincaré embeddings which are trained on the output of handcrafted lexicalsyntactic patterns on a large corpus Anthology ID:  - Information from aclanthology.org

modeling and leveraging fingerprinting techniques to predict a users UF: we retrieve the most frequent opinion from ones history - Information from aclanthology.org

Depression is a global mental health condition that affects all cultures Despite this the way depression is expressed varies by culture Uptake of machine  - Information from aclanthology.org

InfoNCE a kind of mutual information loss that has been applied widely in machine learning and computer vision (van den Oord et al 2018  - Information from aclanthology.org

Nicolas Papernot Patrick McDaniel Ian Goodfellow Somesh Jha Z Berkay Celik and Ananthram Swami 2017 Practical blackbox attacks against machine learning - Information from aclanthology.org

We analyze these negative results to provide insights for future studies Anthology ID: 2020wmt112 Volume: Proceedings of the Fifth Conference on Machine - Information from aclanthology.org

Multilingual BERT (mBERT) trained on 104 languages has shown surprisingly good crosslingual performance on several NLP tasks even without explicit  - Information from aclanthology.org

In this paper we examine two narrativestyle books: Biohazard by Ken Alibek and The Demon in the Freezer by Richard Preston These books are - Information from aclanthology.org

clear understanding of the distinction between form and meaning will help guide the field towards better science around natural language understanding - Information from aclanthology.org

AlexandruLucian Georgescu Horia Cucu Andi Buzo Corneliu Burileanu Language: English URL: https://aclanthologyorg/2020lrec1814 DOI: Bibkey: - Information from aclanthology.org

This paper describes SentencePiece a languageindependent subword tokenizer and detokenizer designed for Neuralbased text processing including Neural  - Information from aclanthology.org

To ground natural language semantics in realworld the generalized GelmanRubin diagnostic method Approaching the symbol grounding problem - Information from aclanthology.org

Training Transformers however requires 18Layer PostLN Transformer training (iethe original the inception architecture for computer vision In - Information from aclanthology.org

We propose a novel network embedding method ie Multiview Adversarially Regularized Graph Autoencoder (MRGAE) - Information from aclanthology.org

Expanding English and Chinese Dictionaries by Wikipedia Titles Often they share moti tic relaxation gibbs distributions and the bayesian - Information from aclanthology.org

Deep Convolutional Neural Network Textual Features and Multiple Kernel Learning for Utterancelevel Multimodal Sentiment Analysis - Information from aclanthology.org

In this paper we present Dialogue Graph Convolutional Network (DialogueGCN) a graph neural network based approach to ERC - Information from aclanthology.org

rule to be consistent across the dictionary Romanian Aromanian on phonotactic rules (Frampton 2011) If the ex - Information from aclanthology.org

meaning via a vector of real values (Deerwester et al 1990) The Word2Vec models introduced by the same choices for optimizing training parameters - Information from aclanthology.org

In 2018 IEEE International Conference on Acoustics Speech and Signal Processing ICASSP 2018 Cal gary AB Canada April 1520 2018 pages  - Information from aclanthology.org

In a corruption of John Searles famous AI thought experiment the Chinese Room (Searle 1980) we twist its original intent by enabling humans to translate  - Information from aclanthology.org

Automatic sociopolitical and crisis event detection has been a challenge for natural language processing as well as social and political science communities  - Information from aclanthology.org

We then survey some dimensions of construed meaning and show how insights from construal could inform theoretical and practical work in NLP - Information from aclanthology.org

Dan Roth Department of Computer Science story comprehension in which the target text is less Based on the SNoW learning architecture we de - Information from aclanthology.org

In this paper we study argument quality assessment in a multimodal context and experiment on DBATES a publicly available dataset of long debate videos First  - Information from aclanthology.org

A Learner Corpusbased Approach to Verb Suggestion for ESL · Yu Sawai Mamoru Komachi Yuji Matsumoto Anthology ID: P132124 Volume: Proceedings of the  - Information from aclanthology.org

Jacob Andreas 2021 pdf bib abs How Do Neural Sequence Models Generalize? Local and Global Cues for Out  - Information from aclanthology.org

Recently neural models led to significant improvements in both machine translation (MT) and natural language generation tasks (NLG) However generation of  - Information from aclanthology.org

with their syntactic trees to improve the grammar of generated sentences Distinct from existing VAEbased text generative models SIVAE contains two  - Information from aclanthology.org

Abstract Communicating complex scientific ideas without misleading or overwhelming the public is challenging While science communication guides exist  - Information from aclanthology.org

the topic distribution of each abstract sentence and decode the sentence from topicaware representations with a PointerGenerator net - Information from aclanthology.org

In this paper we propose a semiautomatic framework for generating disentangled shifts by introducing a controllable visual questionanswer generation (VQAG)  - Information from aclanthology.org

Learning attention functions requires large volumes of data but many NLP tasks simulate human behavior and in this paper we show that human attention really  - Information from aclanthology.org

November 7–11 2021 c 2021 Association for Computational Linguistics 1801 Competency Problems: On Finding and Removing Artifacts in Language Data - Information from aclanthology.org

The Uniqueness of Human Strategies in Simultaneous Interpretation 2Language produced in the process of translation is often con - Information from aclanthology.org

Faculty of Translation Studies Linguistics and Cultural Studies University of Mainz The remaining four translators read the whole text - Information from aclanthology.org

We cast the task as a reinforcement learning problem where we jointly train two systems: a generative model to produce response sequences  - Information from aclanthology.org

First it effectively increases feature interactions by using atrous convolutions Second to address the original information forgotten issue and vanishing/  - Information from aclanthology.org

Thus the sentence is classified as “Others” (ie no blame or praise) 5 Corpus Creation We created our data from the the ISEAR dataset which was collected  - Information from aclanthology.org

Gabriele Pergola Lin Gui Yulan He · Proceedings of the 2021 Conference of the North American Chapter of the Association for Computational Linguistics:  - Information from aclanthology.org

These uncivil interactions can have a powerful negative impact on people by marginalizing minorities and disadvantaged groups The linguistic subtlety of  - Information from aclanthology.org

Many studies are connected to improving OCR ac curacy which remains as an important task for old printed materials Recognizing named entities  - Information from aclanthology.org

Proceedings of the 23rd International Conference on Computational Linguistics Learning Sourceside Reordering Rules for Large Scale Machine Translation - Information from aclanthology.org

BWS Argument Similarity Dataset (BWS): Existing similarity datasets have the disadvantage that the sentence pair selection/sampling process is not always  - Information from aclanthology.org

We describe an online Bayesian change point detection algorithm that we use to detect topic changes from online LDA output Extensive experiments on social  - Information from aclanthology.org

This ranking mechanism have a low attack success rate and is not exhaustive enough to search for adversarial ex amples with low perturbation  - Information from aclanthology.org

pacity and limitations of statistical learning mecha sition gains support from formal learning theory [1 The Nature of Statistical Learning  - Information from aclanthology.org

First it effectively increases feature interactions by using atrous convolutions Second to address the original information forgotten issue and vanishing/  - Information from aclanthology.org

We propose Pretrain and Plugin Variational Auto Encoder (PPVAE) which decouples the text gen eration module from the condition representation module - Information from aclanthology.org

BERTSUM extends BERT by inserting multiple [CLS] symbols to learn sentence representations and using interval segmentation embeddings (illustrated in red and  - Information from aclanthology.org

iii Our statistic analysis of Poincaré embeddings representation of PhySH shows successful hierarchical representation share two characteristics: firstly  - Information from aclanthology.org

2004 Greenwood 2004) This is in contrast to other strategies using a twostep procedure consist ing of document retrieval and searchtime passag - Information from aclanthology.org

Keywords: Automatic thesaurus construction Terminology extraction Semantic similarity 1 Introduction In computational linguistics and terminology  - Information from aclanthology.org

We present a neural network approach to en code relations between sentences in document rep resentation for sentiment classification • We report empirical - Information from aclanthology.org

An autoencoder is an unsupervised learning algorithm that learns to reconstruct its input using a deep neural network Its network structure consists of an  - Information from aclanthology.org

Natural Language Learning pp 640–648 Prague June 2007 cO2007 Association for Computational Linguistics SemiMarkov Models for Sequence Segmentation - Information from aclanthology.org

One of the aspects of social media is the ability for their information producers to hide fully or partially their identity during a discussion  - Information from aclanthology.org

A Convolutional Neural Network for Modelling Sentences · Nal Kalchbrenner Edward Grefenstette Phil Blunsom Anthology ID: P141062 Volume: Proceedings of the  - Information from aclanthology.org

On the Properties of Neural Machine Translation : Encoder–Decoder Approaches Workshop on Syntax Semantics and Structure in Statistical Translation  - Information from aclanthology.org

Generating Sentences from a Continuous Space · Samuel R Bowman Luke Vilnis Oriol Vinyals Andrew Dai Rafal Jozefowicz Samy Bengio  - Information from aclanthology.org

In this paper we present a novel approach to improve the translation performance in NMT by conveying topic knowledge during translation - Information from aclanthology.org

supervised feature learning) and using these features it is trained to discriminate between resourced unsupervised feature learning tech - Information from aclanthology.org

This document describes a typecoherent initial LF or unscoped logical forms (ULF) for EL which captures the predicateargument structure in the EL semantic  - Information from aclanthology.org

heads (NFHs) a linguistic construction that is a In FH constructions the implicit heads are supports this (Section 5) in practice our modeling - Information from aclanthology.org

arXiv preprint arXiv:170307022 Jiasen Lu Jianwei Yang Dhruv Batra and Devi Parikh 2016 Hierarchical questionimage co attention for visual - Information from aclanthology.org

Interactive colorization seeks human input usually in the form of clicks or strokes on the image with malization: Accelerating deep network training by - Information from aclanthology.org

syntactic paradigm the exact meaning of animacy is also a source of variation from one language example in Arabic and Russian animacy has an immediate - Information from aclanthology.org

Unsupervised Data Augmentation (UDA) is a semisupervised technique that applies a consistency loss to penalize differences between a models predictions on  - Information from aclanthology.org

Therefore microblogging websites are rich sources of data for opinion mining and sentiment analysis Because microblogging has appeared relatively recently  - Information from aclanthology.org

Beijing China July 2631 2015 cO2015 Association for Computational Linguistics Harnessing Context Incongruity for Sarcasm Detection Aditya Joshi123 - Information from aclanthology.org

ent text classification datasets across tasks like outperforms other recurrent convolutional ar Convolutional Neural Networks (CNN) and Long - Information from aclanthology.org

Anthology ID: D141181 Volume: Proceedings of the 2014 Conference on Empirical Methods in Natural Language Processing (EMNLP) Month: October Year: 2014  - Information from aclanthology.org

On a casestudy of the negotiation agent developed by (Lewis et al 2017) our attacks reduced the average advantage of rewards between the attacker and the  - Information from aclanthology.org

Convolutional Neural Networks for Sentence Classification Proceedings of the 2014 Conference on Empirical Methods in Natural Language Processing (EMNLP)  - Information from aclanthology.org

In this paper instead of considering Then they apply batch normalization to the variance (2016) utilize batch normalization in all fully con - Information from aclanthology.org

Unsupervised deep embedding for clustering analysis In International conference on machine learning pages 478–487 Omar F Zaidan and Chris CallisonBurch - Information from aclanthology.org

put consistency of data augmentation and apply a selfcritical training strategy (Rennie et al 2017) to exploit large amounts of unsupervised video data - Information from aclanthology.org

of a listeners beliefs capable of predicting commonly drawn inferences We then describe our mechanism for the generation of Peripheral RDs - Information from aclanthology.org

A common instance of selfdisclosure is the start of a conversation with an exchange of names and additional selfintroductions Another example of  - Information from aclanthology.org

document extractive summarization model for work in extractive summarization has explicitly first to apply LSTMminus to text summarization - Information from aclanthology.org

persuasive essays with 1673 sentences and 1552 argumentative units Then they extended their dataset to 402 essays achieving a total of 7116  - Information from aclanthology.org

sequence model Hidden Markov Models (HMMs) Partofspeech tags encode morphosyntactic informa neural networks for handwriting recognition In In - Information from aclanthology.org

We present an approach to unravel antonymy and synonymy from word vectors based on a siamese network inspired approach The model consists of a twophase  - Information from aclanthology.org

Welcome to the ACL Anthology! The ACL Anthology currently hosts 73877 papers on the study of computational linguistics and natural language processing - Information from aclanthology.org

2021 pdf bib abs BagofWords Baselines for Semantic Code Search · Xinyu Zhang Ji Xin Andrew Yates Jimmy Lin - Information from aclanthology.org

However under typical training procedures advantages over classical methods emerge only with large datasets As a result deep learning is employed only when  - Information from aclanthology.org

Speeding up Reinforcement Learningbased Information Extraction Training using Asynchronous Methods asynchronous deep RL methods in IE (and also - Information from aclanthology.org

(2021) used a supervised contrastive loss to improve finetuning performance of pretrained language models in sev eral fewshot learning scenarios In this  - Information from aclanthology.org

Selftraining Improves Pretraining for Natural Language Understanding Jingfei Du†∗ Edouard Grave† Beliz Gunel‡ Vishrav Chaudhary† - Information from aclanthology.org

Interpreting the performance of deep learning models beyond test set accuracy is challenging Characteristics of individual data points are often not  - Information from aclanthology.org

Anthology ID: D131170 Volume: Proceedings of the 2013 Conference on Empirical Methods in Natural Language Processing Month: October Year: 2013 - Information from aclanthology.org

Unmasking Contextual Stereotypes: Measuring and Mitigating BERTs Gender Bias Marion Bartl If for example a model does not associate female terms - Information from aclanthology.org

Our main contributions are summarized here: • We propose MLMLM a mean likelihood method to compare the likelihood of differ ent text of  - Information from aclanthology.org

In this paper we show that commonsense in ference still proves difficult for even state oftheart models by presenting HellaSwag a new challenge dataset - Information from aclanthology.org

Infant directed speech (IDS) is thought to infant could facilitate word segmentation when xxx xxxxxx xxxxxxxx xxx - Information from aclanthology.org

Yucheng Zhou Xiubo Geng Tao Shen Jian Pei Wenqiang Zhang Daxin Jiang · Findings of the Association for Computational Linguistics: ACLIJCNLP 2021 - Information from aclanthology.org

Abstract We present QuAC a dataset for Question Answering in Context that contains 14K informationseeking QA dialogs (100K questions in total) - Information from aclanthology.org

Crossdomain sentiment classification aims to predict sentiment polarity on a target domain utilizing a classifier learned from a source domain Most existing  - Information from aclanthology.org

Enriching bert with knowledge graph embeddings for document classification ArXiv abs/190908402 Michael Schlichtkrull Thomas N Kipf  - Information from aclanthology.org

Mohit Iyyer John Wieting Kevin Gimpel Luke Zettlemoyer generating syntactically adversarial examples that both (1) “fool” pretrained models and (2)  - Information from aclanthology.org

Copy Citation: BibTeX MODS XML Endnote More options… PDF: https://aclanthologyorg/C161282pdf · PDF - Information from aclanthology.org

if they did not serve the cause of a pedagogical e learning system The focus has been to improve Learning through the Internet or elearning paves - Information from aclanthology.org

Anthology ID: W150507 Volume: Proceedings of the 2nd Workshop on Argumentation Mining Month: June Year: 2015 Address: Denver CO Venue: WS SIG:  - Information from aclanthology.org

by JD Finch · 2020 · Cited by 5 — This demo paper presents Emora STDM (State Transition Dialogue Manager) a dialogue system development framework that provides novel workflows for rapid  - Information from aclanthology.org

MultiDomain Named Entity Recognition with GenreAware and Agnostic Inference b) multiple domains where the genre label is unknown at inference time  - Information from aclanthology.org

Extracting rationales can help human understand which information the model utilizes and how it makes the prediction towards better interpretability - Information from aclanthology.org

We propose TEASPN a protocol and an opensource framework for achieving integrated writing assistance environments The protocol standardizes the way  - Information from aclanthology.org

by E Voita · 2019 · Cited by 373 — Multihead selfattention is a key component of the Transformer a stateoftheart architecture for neural machine translation In this work we evaluate  - Information from aclanthology.org

Anthology ID: C962161 Volume: COLING 1996 Volume 2: The 16th International Conference on Computational Linguistics Month: Year: 1996 Address:  - Information from aclanthology.org

Pre editing is the process of modifying the source text (ST) so that it can be translated by machine translation (MT) in a better quality - Information from aclanthology.org

crowdsourcing text relevance sentiments and novelty (1) can be a main step in identifying salient information and (2) provides a deeper - Information from aclanthology.org

Proceedings of the 2020 Conference on Empirical Methods in Natural Language Processing (EMNLP) · Bonnie Webber Trevor Cohn Yulan He Yang Liu (Editors)  - Information from aclanthology.org

I N M c M N N Ln M cc N cc l0 c c lD 0 I M I LD I N M a U r lO m o In n Ln Ln c M O O r r M M r 0 cc N I W m In In m N M M cc c cc M cc N M c b to cc r - Information from aclanthology.org

for Wikipedia articles will help to improve the en averaging gives more weight to smaller categories again existing singlelabel models - Information from aclanthology.org

In this paper we present TaBERT a pretrained LM that jointly learns representations for NL sentences and (semi)structured tables TaBERT is trained on a  - Information from aclanthology.org

Copy Citation: BibTeX MODS XML Endnote More options… PDF: https://aclanthologyorg/E172021pdf · PDF  - Information from aclanthology.org

Inferring User Political Preferences from Streaming Communications Anthology ID: P141018 Volume: Proceedings of the 52nd Annual Meeting of the Association  - Information from aclanthology.org

Jiaqi Li Ming Liu MinYen Kan Zihao Zheng Zekun Wang Wenqiang Lei Ting Liu Bing Qin · Proceedings of the 28th International Conference on  - Information from aclanthology.org

HGCN JE and HGCNJR are our complete models that use joint representations to further improve en tity alignment and relation alignment (Section 44) - Information from aclanthology.org

Peiyu Liu ZeFeng Gao Wayne Xin Zhao ZhiYuan Xie ZhongYi Lu JiRong Wen · Proceedings of the 59th Annual Meeting of the Association for  - Information from aclanthology.org

Anthology ID: W970501 Volume: Natural Language Processing for Communication Aids Month: Year: 1997 Address: Venue: WS SIG: Publisher: Note: Pages:  - Information from aclanthology.org

O 1~ N N N N N I N N ~i M M H N ti M M I N I rl W N M n M N N N N ~ I~j i0 n n T o N N N ~T N al In M C O M V1 1 1 N M M - Information from aclanthology.org

the first time that very deep convolutional nets while the use of (deep) neural networks in NLP troduction of several largescale data sets for text - Information from aclanthology.org

Neural language models (NLMs) which place probability distributions over sequences of words produce contextual word and sentence embeddings - Information from aclanthology.org

Stance classification is a necessary step for inferring these perspectives in terms of supporting or opposing the claim In this work we present a neural  - Information from aclanthology.org

Evaluating and Enhancing the Robustness of Neural Networkbased Dependency Parsing Models with tack towards a specific subtree while the sentence - Information from aclanthology.org

To assist lan guage tutors support novice L2 writers we have also produced tools that help tutors rein force their students independent writing and - Information from aclanthology.org

guess the meaning of a novel word? We present novice learners with new L2 words inserted in sentences otherwise written in their na tive language (L1) - Information from aclanthology.org

nodes and detecting lymph node metastasis in imag (a) Bookmark1: Peripancreatic lymph node the lymph nodes and insert hyperlinks size mea - Information from aclanthology.org

Specifically we construct directed inference graphs for reference captions based on natural language inference A PageRank algorithm is then employed to  - Information from aclanthology.org

Machel Reid Edison MarreseTaylor Yutaka Matsuo · Findings of the Association for Computational Linguistics: EMNLP 2021 - Information from aclanthology.org

The linguistic choices in each variable allow us to study increased rates of acceptances of gay marriage and gender equality respectively In  - Information from aclanthology.org

Distributedness and NonLinearity of LOLITAs Semantic Network 3 SemNet: LOLITAs Semantic ANAl()(1 also provides tits possibility o[ read - Information from aclanthology.org

Representation of text with transferable encoding 1Here we define visual grounding as learning language - Information from aclanthology.org

Sentence representation learning plays a vital role Contrastive Learning for Visual Representa 6https://githubcom/UKPLab/ - Information from aclanthology.org

is through the comprehension of linguistic descrip tions of another persons perceptions and example the concept Tortoise is stereotypically as - Information from aclanthology.org

SimpleQuestions is a commonly used benchmark for singlefactoid question answering (QA) over Knowledge Graphs (KG) Existing QA systems rely on various  - Information from aclanthology.org

We make all of the resources and tools introduced in this study available under open licenses from https://cambridgeltlgithubio/cancerhallmarkcnn/ - Information from aclanthology.org

Dynamic topic models (DTMs) capture the evolution of topics and trends in time series dataCurrent DTMs are applicable only to monolingual datasets - Information from aclanthology.org

Keywords: word embeddings urban dictionary slang sentiment sarcasm ACM Risch J and Krestel R (2019) Domainspecific word - Information from aclanthology.org

Estimating uncertainties of Neural Network predictions paves the way towards more reliable and trustful text classifications However common uncertainty - Information from aclanthology.org

To address the first issue we investigate several types of Long ShortTerm Memory (LSTM) networks that can model both the conversation context and the  - Information from aclanthology.org

Our meaning embedding allows efficient crosslingual sentence similarity estimation by simple cosine similarity calculation Experimental results on both  - Information from aclanthology.org

Coronal Specification and Licensing in Place Assimilation Anthology ID: Y961050 Volume: Proceedings of the 11th Pacific Asia Conference on Language  - Information from aclanthology.org

Vector representations of sentences trained on massive text corpora are widely used as generic sentence embeddings across a variety of NLP problems - Information from aclanthology.org

bines a clustering algorithm with a model of seg with a hierarchical clustering algorithm to derive the They do not require any deep - Information from aclanthology.org

Meta Learning and Its Applications to Natural Language Processing sion rebuttal narrative generation and modeling - Information from aclanthology.org

explore the use of knowledge graphs as a representation for domain knowledge transfer for training textadventure playing reinforcement learning agents - Information from aclanthology.org

With the results of this study we could demonstrate the influence of the users culture and gender the frequency of use of speech based assistants as well as  - Information from aclanthology.org

Normalizing Flow (RNF) for text modeling The RNF transforms a latent variable such that for any two vectors u v in the tangent Wasserstein gan - Information from aclanthology.org

PhraseBased Statistical Machine Translation as a Traveling Salesman Problem · Mikhail Zaslavskiy Marc Dymetman Nicola Cancedda - Information from aclanthology.org

Apart from the interpretability property sparse representations can also improve the usability of word vectors as features The embeddings with good sparsity  - Information from aclanthology.org

In the skipthought model given a sentence tuple (si−1sisi+1) the encoder computes a fixeddimension vector as the representation zi - Information from aclanthology.org

We propose TuckER a relatively straightforward but powerful linear model based on Tucker decomposition of the binary tensor representation of knowledge  - Information from aclanthology.org

of figurative language in Mandarin Chinese – one of which occurs at the Chinese translations were borrowed from the annotations by a  - Information from aclanthology.org

Kevin Yang 2021 pdf bib abs FUDGE: Controlled Text Generation With Future Discriminators · Kevin Yang Dan Klein - Information from aclanthology.org

To evaluate answer ranking features we chose the Academia Sinica Question Answering (ASQA) system as the testbed system for our experiment  - Information from aclanthology.org

sentencelevel autoencoder from a pretrained frozen transformer language model We adapt the masked language modeling objective as a - Information from aclanthology.org

naug is the number of generated augmented sentences per original sentence ing sets overfitting was more likely so generat ing many augmented sentences  - Information from aclanthology.org

tion (PEP) a novel algorithm for approximate inference 22 Inference by (loopy) belief propagation approximate Bayesian inference - Information from aclanthology.org

the medical domain In this paper we propose to generate radiology reports with memory driven Transformer where a relational mem - Information from aclanthology.org

Summaries in XSum are shorter (onesentence) and more abstractive which restricts the semantic diversity of candidates and makes it harder to  - Information from aclanthology.org

Text Generation from Structured Data with Application to the Biography Domain Anthology ID: D161128 Volume: Proceedings of the 2016 Conference on  - Information from aclanthology.org

ranked list returned by WikiRef and the one produced by averaging the responses of Algebric modeling language Bayesian model of computational anatomy  - Information from aclanthology.org

verbal synonym lexicon and corpora with all events annotated by such lexicon labels for one semantic role distinguishing the animacy here (Speaker vs - Information from aclanthology.org

In this paper we propose AdapterDrop removing adapters from lower transformer layers during training and inference which incorporates concepts from all  - Information from aclanthology.org

With the development of deep learning neu a text level graph we connect word nodes within a comprehensive survey on graph neural networks - Information from aclanthology.org

NLP pipelines On the other hand CLAMP does not provide direct support for integrating external tools developed in other programming languages than Java - Information from aclanthology.org

acoustic model trained with multilingual data especially when much larger amount of data from donor (HMM) / Deep Neural Network (DNN) based monolingual - Information from aclanthology.org

In this paper we present TAPAS (for Table Parser) a weakly supervised question answering model that reasons over tables without generating logical forms - Information from aclanthology.org

We study both discriminative models that learn to distinguish coherent from incoherent discourse and generative models that produce coherent text including a  - Information from aclanthology.org

Images and Imagination: Automated Analysis of Priming Effects Related to Autism Spectrum Disorder and Developmental Language Disorder - Information from aclanthology.org

Neural network methods for processing the data represented in graph domains have been studied for several years Graph neural networks (GNNs) have also been  - Information from aclanthology.org

Text classification is a fundamental task in natu ral language processing (NLP) Machine learning and deep learning have achieved high accuracy on tasks  - Information from aclanthology.org

deep model represents a word of such language Extracting and reusing the patterns learnt by a deep neural network (DNN) is a subfield of deep learning - Information from aclanthology.org

It consists of two components: bottomlevel sentence encoders using pretrained transformers and an upperlevel transformerbased segmentation model based on  - Information from aclanthology.org

we propose a planningbased conditional variational autoencoder namely PlanCVAE which first plans a keyword sequence and then generates a story based  - Information from aclanthology.org

Heterogeneous Graph Transformer for GraphtoSequence Learning Lam 2020) are based on Transformer and learn at https://githubcom/QAQv/HetGT - Information from aclanthology.org

African American Vernacular English and Indian English — there are a smaller number of socalled standard Englishes British English and Ameri - Information from aclanthology.org

In SCDV word embeddings are clustered to capture multiple semantic contexts in which words occur They are then chained together to form document topicvectors  - Information from aclanthology.org

As many humor prediction datasets claim to aid in generative tasks Volume: Proceedings of the Second Workshop on Figurative Language Processing  - Information from aclanthology.org

TAPAS concatenates all table cells without adding extra tokens and then ex tends BERTs architecture with additional embed ding layers to  - Information from aclanthology.org

While several researches addressed synonym extraction of SWTs few of them dealt with MWTs and fewer or none while MWTs Dynamic Pooling and Unfold - Information from aclanthology.org

A Grammar of English on Mathematical Principles Zellig Harris John Wiley and Sons New York 1982 Pp xvi 429 $4350 - Information from aclanthology.org

measuring real (vs perceived) productivity gains Kolovratnik (2012) conduct a largescale evaluation of MT PE aimed at estimating the busi - Information from aclanthology.org

Anthology ID: D141166 Volume: Proceedings of the 2014 Conference on Empirical Methods in Natural Language Processing (EMNLP) Month: October Year: 2014  - Information from aclanthology.org

deep convolutional networks which have theart on several public text classification puter vision (eg 1 000 classes for ImageNet) - Information from aclanthology.org

Anthology ID: N151 Month: May–June Year: 2015 Address: Denver Colorado Venue: NAACL  - Information from aclanthology.org

ball slowly rolling towards the edge of a table1 Thus the empirical coverage of the Several psychological theories of narrative under - Information from aclanthology.org

by Adding Automatically Learned Pronunciation Variants to the Lexicon of transcription variants that lie closer to the `true name pronunciations - Information from aclanthology.org

per are used to enable semantic features in at wwwsemanticscholarorg in a step towards before it is submitted for peer review or used to - Information from aclanthology.org

explore the use of knowledge graphs as a representation for domain knowledge transfer for training textadventure playing reinforce ment learning agents - Information from aclanthology.org

to implement the dual structure between language and knowledge in the open domain Then we develop an agent called Orator to accomplish the OIN task  - Information from aclanthology.org

introduce LIDA an annotation tool designed specifically for conversation data As far as we know LIDA is the first dialogue annotation - Information from aclanthology.org

introduces a process that uses a thesaurus to extract cognition results from Finally note that no faults are caused by insufficient training of  - Information from aclanthology.org

theart on several public text classification (eg POS tagging or sentiment analysis) struc while the use of (deep) neural networks in NLP - Information from aclanthology.org

Chinese Buddhist Canon The extracted terms then were filtered through a list of selected transliteration terms from the Buddhist Translation Lexicon and  - Information from aclanthology.org

The problem of learning distributed representations (or embeddings) from a knowledge base (KB) has recently attracted considerable attention These  - Information from aclanthology.org

In classification and regression tasks Bayesian neural networks have been effective in estimating model uncertainty However inference time - Information from aclanthology.org

This paper introduces the novel task of scene segmentation on narrative texts and provides an annotated corpus a discussion of the linguistic and narrative  - Information from aclanthology.org

This paper studies continual learning (CL) of a sequence of aspect sentiment classification (ASC) tasks Although some CL techniques - Information from aclanthology.org

An erratum clarifies errors made in the original scholarly work please be sure to include the Anthology ID of the paper in your email - Information from aclanthology.org

We formulate the task as a multiclass classification problem Given text transcript t(i) and the corresponding label l(i) we need to predict l using t We  - Information from aclanthology.org

Abstract In this paper we study yes/no questions that are naturally occurring — meaning that they are generated in unprompted and unconstrained settings - Information from aclanthology.org

such as documents LSTM need to keep the useful features for a quite long period of is trained with backpropagation and the gradient - Information from aclanthology.org

online communities but also the manner and pat terns of behavior within those discussions (Welser et al 2007) For example a community  - Information from aclanthology.org

BERT MK: Integrating Graph Contextualized Knowledge into Pretrained Language Models Bin He1 Di Zhou1 Jinghui Xiao1 Xin Jiang1 Qun Liu1 Nicholas Jing  - Information from aclanthology.org

A Hierarchical VAE for Calibrating Attributes while Generating Text using Normalizing Flow Bidisha Samanta IIT Kharagpur - Information from aclanthology.org

uted training and Transformer While we focus on hybrid datamodel parallelism for Seq2Seq RNN MT in this paper Wang et al - Information from aclanthology.org

The increasing prevalence of political bias in news media calls for greater public awareness of it as well as robust methods for its detection - Information from aclanthology.org

Domain specific words (DSWГs) are important the same word sense if they belong to the same common domain the following DSW (DomainSpecific Word) - Information from aclanthology.org

Experiments related in a first part to distress detection in realtime using audio and speech analysis and in a second part to fall detection using video - Information from aclanthology.org

RL2 : Fast reinforcement learning via slow reinforcement learn ing arXiv:161102779 Chelsea Finn Pieter Abbeel and Sergey Levine 2017 - Information from aclanthology.org

in Naturally Occurring Multimodal Texts Ronja Utescher some examples of realworld multimodal texts ie - Information from aclanthology.org

In this paper we present TaBERT a pretrained LM that jointly learns representations for NL sentences and (semi)structured tables TaBERT is trained on a  - Information from aclanthology.org

In this work we propose to transform typical dense word vectors into disentangled embeddings featuring improved interpretability via encoding polysemous  - Information from aclanthology.org

(2010) is concerned with English learner texts assuming only standard POS labels at three layers: distribution (syntax) morphology and lexical  - Information from aclanthology.org

Variational Neural Machine Translation with Normalizing Flows Hendra Setiawan As we adopt the amortized inference strategy https://github - Information from aclanthology.org

North American Chapter of the Association for Computational Linguistics (2021) Sentence Comprehension Deficits in Aphasia: A Computational Evaluation of  - Information from aclanthology.org

Proceedings of the Workshop on Automated Extraction of Sociopolitical Events Improving Intent Classification in an Ecommerce Voice Assistant by Using  - Information from aclanthology.org

Our generated medical reports on the other hand are fluent and more importantly clinically accurate This is achieved by our fully differentiable and endto  - Information from aclanthology.org

support obscure or reveal the properties of language and therefore affect what Relevant linguistic generalizations are based on speech not writing - Information from aclanthology.org

A Consistency Training Framework for Conversational Question Answering Gangwoo Kim Hyunjae Kim Jungsoo Park Jaewoo Kang† Korea University - Information from aclanthology.org

While the DisCoCat model (Coecke et al 2010) has been proved a valuable tool for studying compositional aspects of language at the level of semantics  - Information from aclanthology.org

Largescale language models such as GPT 3 are excellent fewshot learners allowing them to be controlled via natural text prompts - Information from aclanthology.org

explicitly collecting new data It has received active attention in recent machine learning (ML) research in the form of wellreceived  - Information from aclanthology.org

We report on a series of experiments with convolutional neural networks (CNN) trained on top of pretrained word vec tors for sentencelevel classification - Information from aclanthology.org

Our experiments on entity typing with finegrained classes demonstrate the effectiveness of multiview learning We release MVET a large multiview — and  - Information from aclanthology.org

School of Computing and Digital Technology Birmingham City University UK hansihettiarachchimailbcuacuk {mariamadedoyinolowejagdevbhogal  - Information from aclanthology.org

However at test time the word choice is affected by unknown sources thus “not” may not be associated with the label “contradiction” A model relying on the  - Information from aclanthology.org

Qi “Feature Squeezing: Detecting Adversarial Examples in Deep Neural Networks” 2018 Network and Distributed System Security Symposium (NDSS18) Feb 2018 - Information from aclanthology.org

In this paper we extend recent advances in emoji prediction by putting forward a multimodal approach that is able to predict emojis in Instagram posts - Information from aclanthology.org

This paper describes an extension to the hidden Markov A hidden Markov model (HMM) is a statistical There are many possible answers some of which - Information from aclanthology.org

(UDv2) treebank for Serbian starting from an existing Croatian UDv1 treebank and taking into account the other Slavic UD annotation guidelines - Information from aclanthology.org

We describe a datadriven approach for automatically explaining new nonstandard English expressions in a given sentence building on a large dataset that  - Information from aclanthology.org

of improving the efficiency of dialog policy learn ing (namely user modeling hindsight experience replay and reward shaping) and qualitatively com - Information from aclanthology.org

M M M M 00 CO In N M tD In I b CO O 1 C N N 1 N N N N N I I M ~ ~ N I 1 W 01 U I I M O 1 M M M CO CO CO CO C O) to - Information from aclanthology.org

We propose a novel loss function called hierarchical consistency loss to maintain consistency among the attention weights of the decoders To evaluate the  - Information from aclanthology.org

Proceedings of the 6th Workshop on Representation Learning for NLP (RepL4NLP2021) pages 316–321 Training deep nets with sublinear memory cost - Information from aclanthology.org

CodeBERT learns generalpurpose representations that support downstream NLPL applications such as natural language code search code documentation  - Information from aclanthology.org

the Information Retrieval with Deep Semantic Matching and Deep Rank It has been applied to the online question answer system for insurance Question Answer - Information from aclanthology.org

model with multihead selfattention can be in beling model based on multihead selfattention to 3https://githubcom/googleresearch/ - Information from aclanthology.org

The first sequence tosequence approach is using only the last dialogue turn as a context 32 Dialogue Datasets Feature ConvAI1 ConvAI2 # Dialogues - Information from aclanthology.org

Figure 1: The Architecture of Capsule network for text classification The processes of dynamic routing between consecutive layers are shown in the bottom - Information from aclanthology.org

In regular languages the local context model thus hypothesizes that lexical information governs outofdistribution prediction predicting that  - Information from aclanthology.org

Then the learned knowledge is integrated with a pretrained language model to do the knowledge generalization Experimental results demonstrate that our  - Information from aclanthology.org

tensive collection of Java and Python functions CodeBERT (Feng et al 2020) and GraphCode ation translation program repair clone detection - Information from aclanthology.org

lingual language model pretraining as maximizing mutual information between multilingualmultigranularity texts The unified view helps us to better  - Information from aclanthology.org

compared with projected links extracted from a technical thesaurus 1 Motivation In the domain of corpusbased terminology - Information from aclanthology.org

TaBERT (Yin et al 2020) and TaPas (Herzig et al 2https://githubcom/SFIG611/tabbie TaBERT and TaPaS contextualize cell embeddings - Information from aclanthology.org

tual QA system uses Information Retrieval as doc ument/passage prefetch The ensuing Answer Ex traction operation assumes that one of the passages - Information from aclanthology.org

Abstract A major aim of opensource NLP is to quickly and accurately reproduce the results of new work in a manner - Information from aclanthology.org