paper_id
stringlengths
15
18
title
stringlengths
22
143
paper_url
stringlengths
41
44
authors
listlengths
1
16
abstract
large_stringlengths
439
1.79k
anthology_id
stringlengths
15
18
doi
stringlengths
27
30
award
stringclasses
3 values
arxiv_id
stringlengths
10
10
arxiv_id_source
stringclasses
2 values
2021.acl-long.401
Data Augmentation with Adversarial Training for Cross-Lingual NLI
https://aclanthology.org/2021.acl-long.401/
[ "Xin Dong", "Yaxin Zhu", "Zuohui Fu", "Dongkuan Xu", "Gerard de Melo" ]
Due to recent pretrained multilingual representation models, it has become feasible to exploit labeled data from one language to train a cross-lingual model that can then be applied to multiple new languages. In practice, however, we still face the problem of scarce labeled data, leading to subpar results. In this pape...
2021.acl-long.401
10.18653/v1/2021.acl-long.401
null
null
null
2021.acl-long.402
Bootstrapped Unsupervised Sentence Representation Learning
https://aclanthology.org/2021.acl-long.402/
[ "Yan Zhang", "Ruidan He", "Zuozhu Liu", "Lidong Bing", "Haizhou Li" ]
As high-quality labeled data is scarce, unsupervised sentence representation learning has attracted much attention. In this paper, we propose a new framework with a two-branch Siamese Network which maximizes the similarity between two augmented views of each sentence. Specifically, given one augmented view of the input...
2021.acl-long.402
10.18653/v1/2021.acl-long.402
null
null
null
2021.acl-long.403
Learning Event Graph Knowledge for Abductive Reasoning
https://aclanthology.org/2021.acl-long.403/
[ "Li Du", "Xiao Ding", "Ting Liu", "Bing Qin" ]
Abductive reasoning aims at inferring the most plausible explanation for observed events, which would play critical roles in various NLP applications, such as reading comprehension and question answering. To facilitate this task, a narrative text based abductive reasoning task \alphaNLI is proposed, together with explo...
2021.acl-long.403
10.18653/v1/2021.acl-long.403
null
null
null
2021.acl-long.404
A Cognitive Regularizer for Language Modeling
https://aclanthology.org/2021.acl-long.404/
[ "Jason Wei", "Clara Meister", "Ryan Cotterell" ]
The uniform information density (UID) hypothesis, which posits that speakers behaving optimally tend to distribute information uniformly across a linguistic signal, has gained traction in psycholinguistics as an explanation for certain syntactic, morphological, and prosodic choices. In this work, we explore whether the...
2021.acl-long.404
10.18653/v1/2021.acl-long.404
null
2105.07144
title_snapshot
2021.acl-long.405
Lower Perplexity is Not Always Human-Like
https://aclanthology.org/2021.acl-long.405/
[ "Tatsuki Kuribayashi", "Yohei Oseki", "Takumi Ito", "Ryo Yoshida", "Masayuki Asahara", "Kentaro Inui" ]
In computational psycholinguistics, various language models have been evaluated against human reading behavior (e.g., eye movement) to build human-like computational models. However, most previous efforts have focused almost exclusively on English, despite the recent trend towards linguistic universal within the genera...
2021.acl-long.405
10.18653/v1/2021.acl-long.405
null
2106.01229
title_snapshot
2021.acl-long.406
Word Sense Disambiguation: Towards Interactive Context Exploitation from Both Word and Sense Perspectives
https://aclanthology.org/2021.acl-long.406/
[ "Ming Wang", "Yinglin Wang" ]
Lately proposed Word Sense Disambiguation (WSD) systems have approached the estimated upper bound of the task on standard evaluation benchmarks. However, these systems typically implement the disambiguation of words in a document almost independently, underutilizing sense and word dependency in context. In this paper, ...
2021.acl-long.406
10.18653/v1/2021.acl-long.406
null
null
null
2021.acl-long.407
A Knowledge-Guided Framework for Frame Identification
https://aclanthology.org/2021.acl-long.407/
[ "Xuefeng Su", "Ru Li", "Xiaoli Li", "Jeff Z. Pan", "Hu Zhang", "Qinghua Chai", "Xiaoqi Han" ]
Frame Identification (FI) is a fundamental and challenging task in frame semantic parsing. The task aims to find the exact frame evoked by a target word in a given sentence. It is generally regarded as a classification task in existing work, where frames are treated as discrete labels or represented using onehot embedd...
2021.acl-long.407
10.18653/v1/2021.acl-long.407
null
null
null
2021.acl-long.408
Obtaining Better Static Word Embeddings Using Contextual Embedding Models
https://aclanthology.org/2021.acl-long.408/
[ "Prakhar Gupta", "Martin Jaggi" ]
The advent of contextual word embeddings — representations of words which incorporate semantic and syntactic information from their context—has led to tremendous improvements on a wide variety of NLP tasks. However, recent contextual models have prohibitively high computational cost in many use-cases and are often hard...
2021.acl-long.408
10.18653/v1/2021.acl-long.408
null
2106.04302
title_snapshot
2021.acl-long.409
Meta-Learning with Variational Semantic Memory for Word Sense Disambiguation
https://aclanthology.org/2021.acl-long.409/
[ "Yingjun Du", "Nithin Holla", "Xiantong Zhen", "Cees Snoek", "Ekaterina Shutova" ]
A critical challenge faced by supervised word sense disambiguation (WSD) is the lack of large annotated datasets with sufficient coverage of words in their diversity of senses. This inspired recent research on few-shot WSD using meta-learning. While such work has successfully applied meta-learning to learn new word sen...
2021.acl-long.409
10.18653/v1/2021.acl-long.409
null
2106.02960
title_snapshot
2021.acl-long.410
LexFit: Lexical Fine-Tuning of Pretrained Language Models
https://aclanthology.org/2021.acl-long.410/
[ "Ivan Vulić", "Edoardo Maria Ponti", "Anna Korhonen", "Goran Glavaš" ]
Transformer-based language models (LMs) pretrained on large text collections implicitly store a wealth of lexical semantic knowledge, but it is non-trivial to extract that knowledge effectively from their parameters. Inspired by prior work on semantic specialization of static word embedding (WE) models, we show that it...
2021.acl-long.410
10.18653/v1/2021.acl-long.410
null
null
null
2021.acl-long.411
Text-Free Image-to-Speech Synthesis Using Learned Segmental Units
https://aclanthology.org/2021.acl-long.411/
[ "Wei-Ning Hsu", "David Harwath", "Tyler Miller", "Christopher Song", "James Glass" ]
In this paper we present the first model for directly synthesizing fluent, natural-sounding spoken audio captions for images that does not require natural language text as an intermediate representation or source of supervision. Instead, we connect the image captioning module and the speech synthesis module with a set ...
2021.acl-long.411
10.18653/v1/2021.acl-long.411
null
2012.15454
title_snapshot
2021.acl-long.412
CTFN: Hierarchical Learning for Multimodal Sentiment Analysis Using Coupled-Translation Fusion Network
https://aclanthology.org/2021.acl-long.412/
[ "Jiajia Tang", "Kang Li", "Xuanyu Jin", "Andrzej Cichocki", "Qibin Zhao", "Wanzeng Kong" ]
Multimodal sentiment analysis is the challenging research area that attends to the fusion of multiple heterogeneous modalities. The main challenge is the occurrence of some missing modalities during the multimodal fusion procedure. However, the existing techniques require all modalities as input, thus are sensitive to ...
2021.acl-long.412
10.18653/v1/2021.acl-long.412
null
null
null
2021.acl-long.413
Positional Artefacts Propagate Through Masked Language Model Embeddings
https://aclanthology.org/2021.acl-long.413/
[ "Ziyang Luo", "Artur Kulmizev", "Xiaoxi Mao" ]
In this work, we demonstrate that the contextualized word vectors derived from pretrained masked language model-based encoders share a common, perhaps undesirable pattern across layers. Namely, we find cases of persistent outlier neurons within BERT and RoBERTa’s hidden state vectors that consistently bear the smallest...
2021.acl-long.413
10.18653/v1/2021.acl-long.413
null
2011.04393
title_snapshot
2021.acl-long.414
Language Model Evaluation Beyond Perplexity
https://aclanthology.org/2021.acl-long.414/
[ "Clara Meister", "Ryan Cotterell" ]
We propose an alternate approach to quantifying how well language models learn natural language: we ask how well they match the statistical tendencies of natural language. To answer this question, we analyze whether text generated from language models exhibits the statistical tendencies present in the human-generated t...
2021.acl-long.414
10.18653/v1/2021.acl-long.414
null
2106.00085
title_snapshot
2021.acl-long.415
Learning to Explain: Generating Stable Explanations Fast
https://aclanthology.org/2021.acl-long.415/
[ "Xuelin Situ", "Ingrid Zukerman", "Cecile Paris", "Sameen Maruf", "Gholamreza Haffari" ]
The importance of explaining the outcome of a machine learning model, especially a black-box model, is widely acknowledged. Recent approaches explain an outcome by identifying the contributions of input features to this outcome. In environments involving large black-box models or complex inputs, this leads to computati...
2021.acl-long.415
10.18653/v1/2021.acl-long.415
null
null
null
2021.acl-long.416
StereoSet: Measuring stereotypical bias in pretrained language models
https://aclanthology.org/2021.acl-long.416/
[ "Moin Nadeem", "Anna Bethke", "Siva Reddy" ]
A stereotype is an over-generalized belief about a particular group of people, e.g., Asians are good at math or African Americans are athletic. Such beliefs (biases) are known to hurt target groups. Since pretrained language models are trained on large real-world data, they are known to capture stereotypical biases. It...
2021.acl-long.416
10.18653/v1/2021.acl-long.416
null
2004.09456
title_snapshot
2021.acl-long.417
Alignment Rationale for Natural Language Inference
https://aclanthology.org/2021.acl-long.417/
[ "Zhongtao Jiang", "Yuanzhe Zhang", "Zhao Yang", "Jun Zhao", "Kang Liu" ]
Deep learning models have achieved great success on the task of Natural Language Inference (NLI), though only a few attempts try to explain their behaviors. Existing explanation methods usually pick prominent features such as words or phrases from the input text. However, for NLI, alignments among words or phrases are ...
2021.acl-long.417
10.18653/v1/2021.acl-long.417
null
null
null
2021.acl-long.418
Enabling Lightweight Fine-tuning for Pre-trained Language Model Compression based on Matrix Product Operators
https://aclanthology.org/2021.acl-long.418/
[ "Peiyu Liu", "Ze-Feng Gao", "Wayne Xin Zhao", "Zhi-Yuan Xie", "Zhong-Yi Lu", "Ji-Rong Wen" ]
This paper presents a novel pre-trained language models (PLM) compression approach based on the matrix product operator (short as MPO) from quantum many-body physics. It can decompose an original matrix into central tensors (containing the core information) and auxiliary tensors (with only a small proportion of paramet...
2021.acl-long.418
10.18653/v1/2021.acl-long.418
null
2106.02205
title_snapshot
2021.acl-long.419
On Sample Based Explanation Methods for NLP: Faithfulness, Efficiency and Semantic Evaluation
https://aclanthology.org/2021.acl-long.419/
[ "Wei Zhang", "Ziming Huang", "Yada Zhu", "Guangnan Ye", "Xiaodong Cui", "Fan Zhang" ]
In the recent advances of natural language processing, the scale of the state-of-the-art models and datasets is usually extensive, which challenges the application of sample-based explanation methods in many aspects, such as explanation interpretability, efficiency, and faithfulness. In this work, for the first time, w...
2021.acl-long.419
10.18653/v1/2021.acl-long.419
null
2106.04753
title_judge
2021.acl-long.420
Syntax-Enhanced Pre-trained Model
https://aclanthology.org/2021.acl-long.420/
[ "Zenan Xu", "Daya Guo", "Duyu Tang", "Qinliang Su", "Linjun Shou", "Ming Gong", "Wanjun Zhong", "Xiaojun Quan", "Daxin Jiang", "Nan Duan" ]
We study the problem of leveraging the syntactic structure of text to enhance pre-trained models such as BERT and RoBERTa. Existing methods utilize syntax of text either in the pre-training stage or in the fine-tuning stage, so that they suffer from discrepancy between the two stages. Such a problem would lead to the n...
2021.acl-long.420
10.18653/v1/2021.acl-long.420
null
2012.14116
title_snapshot
2021.acl-long.421
Matching Distributions between Model and Data: Cross-domain Knowledge Distillation for Unsupervised Domain Adaptation
https://aclanthology.org/2021.acl-long.421/
[ "Bo Zhang", "Xiaoming Zhang", "Yun Liu", "Lei Cheng", "Zhoujun Li" ]
Unsupervised Domain Adaptation (UDA) aims to transfer the knowledge of source domain to the unlabeled target domain. Existing methods typically require to learn to adapt the target model by exploiting the source data and sharing the network architecture across domains. However, this pipeline makes the source data risky...
2021.acl-long.421
10.18653/v1/2021.acl-long.421
null
null
null
2021.acl-long.422
Counterfactual Inference for Text Classification Debiasing
https://aclanthology.org/2021.acl-long.422/
[ "Chen Qian", "Fuli Feng", "Lijie Wen", "Chunping Ma", "Pengjun Xie" ]
Today’s text classifiers inevitably suffer from unintended dataset biases, especially the document-level label bias and word-level keyword bias, which may hurt models’ generalization. Many previous studies employed data-level manipulations or model-level balancing mechanisms to recover unbiased distributions and thus p...
2021.acl-long.422
10.18653/v1/2021.acl-long.422
null
null
null
2021.acl-long.423
HieRec: Hierarchical User Interest Modeling for Personalized News Recommendation
https://aclanthology.org/2021.acl-long.423/
[ "Tao Qi", "Fangzhao Wu", "Chuhan Wu", "Peiru Yang", "Yang Yu", "Xing Xie", "Yongfeng Huang" ]
User interest modeling is critical for personalized news recommendation. Existing news recommendation methods usually learn a single user embedding for each user from their previous behaviors to represent their overall interest. However, user interest is usually diverse and multi-grained, which is difficult to be accur...
2021.acl-long.423
10.18653/v1/2021.acl-long.423
null
2106.04408
title_snapshot
2021.acl-long.424
PP-Rec: News Recommendation with Personalized User Interest and Time-aware News Popularity
https://aclanthology.org/2021.acl-long.424/
[ "Tao Qi", "Fangzhao Wu", "Chuhan Wu", "Yongfeng Huang" ]
Personalized news recommendation methods are widely used in online news services. These methods usually recommend news based on the matching between news content and user interest inferred from historical behaviors. However, these methods usually have difficulties in making accurate recommendations to cold-start users,...
2021.acl-long.424
10.18653/v1/2021.acl-long.424
null
2106.01300
title_snapshot
2021.acl-long.425
Article Reranking by Memory-Enhanced Key Sentence Matching for Detecting Previously Fact-Checked Claims
https://aclanthology.org/2021.acl-long.425/
[ "Qiang Sheng", "Juan Cao", "Xueyao Zhang", "Xirong Li", "Lei Zhong" ]
False claims that have been previously fact-checked can still spread on social media. To mitigate their continual spread, detecting previously fact-checked claims is indispensable. Given a claim, existing works focus on providing evidence for detection by reranking candidate fact-checking articles (FC-articles) retriev...
2021.acl-long.425
10.18653/v1/2021.acl-long.425
null
2112.10322
title_snapshot
2021.acl-long.426
Defense against Synonym Substitution-based Adversarial Attacks via Dirichlet Neighborhood Ensemble
https://aclanthology.org/2021.acl-long.426/
[ "Yi Zhou", "Xiaoqing Zheng", "Cho-Jui Hsieh", "Kai-Wei Chang", "Xuanjing Huang" ]
Although deep neural networks have achieved prominent performance on many NLP tasks, they are vulnerable to adversarial examples. We propose Dirichlet Neighborhood Ensemble (DNE), a randomized method for training a robust model to defense synonym substitution-based attacks. During training, DNE forms virtual sentences ...
2021.acl-long.426
10.18653/v1/2021.acl-long.426
null
2006.11627
title_judge
2021.acl-long.427
Shortformer: Better Language Modeling using Shorter Inputs
https://aclanthology.org/2021.acl-long.427/
[ "Ofir Press", "Noah A. Smith", "Mike Lewis" ]
Increasing the input length has been a driver of progress in language modeling with transformers. We identify conditions where shorter inputs are not harmful, and achieve perplexity and efficiency improvements through two new methods that decrease input length. First, we show that initially training a model on short su...
2021.acl-long.427
10.18653/v1/2021.acl-long.427
null
2012.15832
title_snapshot
2021.acl-long.428
BanditMTL: Bandit-based Multi-task Learning for Text Classification
https://aclanthology.org/2021.acl-long.428/
[ "Yuren Mao", "Zekai Wang", "Weiwei Liu", "Xuemin Lin", "Wenbin Hu" ]
Task variance regularization, which can be used to improve the generalization of Multi-task Learning (MTL) models, remains unexplored in multi-task text classification. Accordingly, to fill this gap, this paper investigates how the task might be effectively regularized, and consequently proposes a multi-task learning m...
2021.acl-long.428
10.18653/v1/2021.acl-long.428
null
null
null
2021.acl-long.429
Unified Interpretation of Softmax Cross-Entropy and Negative Sampling: With Case Study for Knowledge Graph Embedding
https://aclanthology.org/2021.acl-long.429/
[ "Hidetaka Kamigaito", "Katsuhiko Hayashi" ]
In knowledge graph embedding, the theoretical relationship between the softmax cross-entropy and negative sampling loss functions has not been investigated. This makes it difficult to fairly compare the results of the two different loss functions. We attempted to solve this problem by using the Bregman divergence to pr...
2021.acl-long.429
10.18653/v1/2021.acl-long.429
null
2106.07250
title_snapshot
2021.acl-long.430
De-Confounded Variational Encoder-Decoder for Logical Table-to-Text Generation
https://aclanthology.org/2021.acl-long.430/
[ "Wenqing Chen", "Jidong Tian", "Yitian Li", "Hao He", "Yaohui Jin" ]
Logical table-to-text generation aims to automatically generate fluent and logically faithful text from tables. The task remains challenging where deep learning models often generated linguistically fluent but logically inconsistent text. The underlying reason may be that deep learning models often capture surface-leve...
2021.acl-long.430
10.18653/v1/2021.acl-long.430
null
null
null
2021.acl-long.431
Rethinking Stealthiness of Backdoor Attack against NLP Models
https://aclanthology.org/2021.acl-long.431/
[ "Wenkai Yang", "Yankai Lin", "Peng Li", "Jie Zhou", "Xu Sun" ]
Recent researches have shown that large natural language processing (NLP) models are vulnerable to a kind of security threat called the Backdoor Attack. Backdoor attacked models can achieve good performance on clean test sets but perform badly on those input sentences injected with designed trigger words. In this work,...
2021.acl-long.431
10.18653/v1/2021.acl-long.431
null
null
null
2021.acl-long.432
Crowdsourcing Learning as Domain Adaptation: A Case Study on Named Entity Recognition
https://aclanthology.org/2021.acl-long.432/
[ "Xin Zhang", "Guangwei Xu", "Yueheng Sun", "Meishan Zhang", "Pengjun Xie" ]
Crowdsourcing is regarded as one prospective solution for effective supervised learning, aiming to build large-scale annotated training data by crowd workers. Previous studies focus on reducing the influences from the noises of the crowdsourced annotations for supervised models. We take a different point in this work, ...
2021.acl-long.432
10.18653/v1/2021.acl-long.432
null
2105.14980
title_snapshot
2021.acl-long.433
Exploring Distantly-Labeled Rationales in Neural Network Models
https://aclanthology.org/2021.acl-long.433/
[ "Quzhe Huang", "Shengqi Zhu", "Yansong Feng", "Dongyan Zhao" ]
Recent studies strive to incorporate various human rationales into neural networks to improve model performance, but few pay attention to the quality of the rationales. Most existing methods distribute their models’ focus to distantly-labeled rationale words entirely and equally, while ignoring the potential important ...
2021.acl-long.433
10.18653/v1/2021.acl-long.433
null
2106.01809
title_snapshot
2021.acl-long.434
Learning to Perturb Word Embeddings for Out-of-distribution QA
https://aclanthology.org/2021.acl-long.434/
[ "Seanie Lee", "Minki Kang", "Juho Lee", "Sung Ju Hwang" ]
QA models based on pretrained language models have achieved remarkable performance on various benchmark datasets. However, QA models do not generalize well to unseen data that falls outside the training distribution, due to distributional shifts. Data augmentation (DA) techniques which drop/replace words have shown to ...
2021.acl-long.434
10.18653/v1/2021.acl-long.434
null
2105.02692
title_snapshot
2021.acl-long.435
Maria: A Visual Experience Powered Conversational Agent
https://aclanthology.org/2021.acl-long.435/
[ "Zujie Liang", "Huang Hu", "Can Xu", "Chongyang Tao", "Xiubo Geng", "Yining Chen", "Fan Liang", "Daxin Jiang" ]
Arguably, the visual perception of conversational agents to the physical world is a key way for them to exhibit the human-like intelligence. Image-grounded conversation is thus proposed to address this challenge. Existing works focus on exploring the multimodal dialog models that ground the conversation on a given imag...
2021.acl-long.435
10.18653/v1/2021.acl-long.435
null
2105.13073
title_snapshot
2021.acl-long.436
A Human-machine Collaborative Framework for Evaluating Malevolence in Dialogues
https://aclanthology.org/2021.acl-long.436/
[ "Yangjun Zhang", "Pengjie Ren", "Maarten de Rijke" ]
Conversational dialogue systems (CDSs) are hard to evaluate due to the complexity of natural language. Automatic evaluation of dialogues often shows insufficient correlation with human judgements. Human evaluation is reliable but labor-intensive. We introduce a human-machine collaborative framework, HMCEval, that can g...
2021.acl-long.436
10.18653/v1/2021.acl-long.436
null
null
null
2021.acl-long.437
Generating Relevant and Coherent Dialogue Responses using Self-Separated Conditional Variational AutoEncoders
https://aclanthology.org/2021.acl-long.437/
[ "Bin Sun", "Shaoxiong Feng", "Yiwei Li", "Jiamou Liu", "Kan Li" ]
Conditional Variational AutoEncoder (CVAE) effectively increases the diversity and informativeness of responses in open-ended dialogue generation tasks through enriching the context vector with sampled latent variables. However, due to the inherent one-to-many and many-to-one phenomena in human dialogues, the sampled l...
2021.acl-long.437
10.18653/v1/2021.acl-long.437
null
2106.03410
title_snapshot
2021.acl-long.438
Learning to Ask Conversational Questions by Optimizing Levenshtein Distance
https://aclanthology.org/2021.acl-long.438/
[ "Zhongkun Liu", "Pengjie Ren", "Zhumin Chen", "Zhaochun Ren", "Maarten de Rijke", "Ming Zhou" ]
Conversational Question Simplification (CQS) aims to simplify self-contained questions into conversational ones by incorporating some conversational characteristics, e.g., anaphora and ellipsis. Existing maximum likelihood estimation based methods often get trapped in easily learned tokens as all tokens are treated equ...
2021.acl-long.438
10.18653/v1/2021.acl-long.438
null
2106.15903
title_snapshot
2021.acl-long.439
DVD: A Diagnostic Dataset for Multi-step Reasoning in Video Grounded Dialogue
https://aclanthology.org/2021.acl-long.439/
[ "Hung Le", "Chinnadhurai Sankar", "Seungwhan Moon", "Ahmad Beirami", "Alborz Geramifard", "Satwik Kottur" ]
A video-grounded dialogue system is required to understand both dialogue, which contains semantic dependencies from turn to turn, and video, which contains visual cues of spatial and temporal scene variations. Building such dialogue systems is a challenging problem, involving various reasoning types on both visual and ...
2021.acl-long.439
10.18653/v1/2021.acl-long.439
null
2101.00151
title_snapshot
2021.acl-long.440
MMGCN: Multimodal Fusion via Deep Graph Convolution Network for Emotion Recognition in Conversation
https://aclanthology.org/2021.acl-long.440/
[ "Jingwen Hu", "Yuchen Liu", "Jinming Zhao", "Qin Jin" ]
Emotion recognition in conversation (ERC) is a crucial component in affective dialogue systems, which helps the system understand users’ emotions and generate empathetic responses. However, most works focus on modeling speaker and contextual information primarily on the textual modality or simply leveraging multimodal ...
2021.acl-long.440
10.18653/v1/2021.acl-long.440
null
2107.06779
title_snapshot
2021.acl-long.441
DynaEval: Unifying Turn and Dialogue Level Evaluation
https://aclanthology.org/2021.acl-long.441/
[ "Chen Zhang", "Yiming Chen", "Luis Fernando D’Haro", "Yan Zhang", "Thomas Friedrichs", "Grandee Lee", "Haizhou Li" ]
A dialogue is essentially a multi-turn interaction among interlocutors. Effective evaluation metrics should reflect the dynamics of such interaction. Existing automatic metrics are focused very much on the turn-level quality, while ignoring such dynamics. To this end, we propose DynaEval, a unified automatic evaluation...
2021.acl-long.441
10.18653/v1/2021.acl-long.441
null
2106.01112
title_snapshot
2021.acl-long.442
CoSQA: 20,000+ Web Queries for Code Search and Question Answering
https://aclanthology.org/2021.acl-long.442/
[ "Junjie Huang", "Duyu Tang", "Linjun Shou", "Ming Gong", "Ke Xu", "Daxin Jiang", "Ming Zhou", "Nan Duan" ]
Finding codes given natural language query is beneficial to the productivity of software developers. Future progress towards better semantic matching between query and code requires richer supervised training resources. To remedy this, we introduce CoSQA dataset. It includes 20,604 labels for pairs of natural language ...
2021.acl-long.442
10.18653/v1/2021.acl-long.442
null
2105.13239
title_snapshot
2021.acl-long.443
Rewriter-Evaluator Architecture for Neural Machine Translation
https://aclanthology.org/2021.acl-long.443/
[ "Yangming Li", "Kaisheng Yao" ]
A few approaches have been developed to improve neural machine translation (NMT) models with multiple passes of decoding. However, their performance gains are limited because of lacking proper policies to terminate the multi-pass process. To address this issue, we introduce a novel architecture of Rewriter-Evaluator. T...
2021.acl-long.443
10.18653/v1/2021.acl-long.443
null
2012.05414
title_snapshot
2021.acl-long.444
Modeling Bilingual Conversational Characteristics for Neural Chat Translation
https://aclanthology.org/2021.acl-long.444/
[ "Yunlong Liang", "Fandong Meng", "Yufeng Chen", "Jinan Xu", "Jie Zhou" ]
Neural chat translation aims to translate bilingual conversational text, which has a broad application in international exchanges and cooperation. Despite the impressive performance of sentence-level and context-aware Neural Machine Translation (NMT), there still remain challenges to translate bilingual conversational ...
2021.acl-long.444
10.18653/v1/2021.acl-long.444
null
2107.11164
title_snapshot
2021.acl-long.445
Importance-based Neuron Allocation for Multilingual Neural Machine Translation
https://aclanthology.org/2021.acl-long.445/
[ "Wanying Xie", "Yang Feng", "Shuhao Gu", "Dong Yu" ]
Multilingual neural machine translation with a single model has drawn much attention due to its capability to deal with multiple languages. However, the current multilingual translation paradigm often makes the model tend to preserve the general knowledge, but ignore the language-specific knowledge. Some previous works...
2021.acl-long.445
10.18653/v1/2021.acl-long.445
null
2107.06569
title_snapshot
2021.acl-long.446
Transfer Learning for Sequence Generation: from Single-source to Multi-source
https://aclanthology.org/2021.acl-long.446/
[ "Xuancheng Huang", "Jingfang Xu", "Maosong Sun", "Yang Liu" ]
Multi-source sequence generation (MSG) is an important kind of sequence generation tasks that takes multiple sources, including automatic post-editing, multi-source translation, multi-document summarization, etc. As MSG tasks suffer from the data scarcity problem and recent pretrained models have been proven to be effe...
2021.acl-long.446
10.18653/v1/2021.acl-long.446
null
2105.14809
title_snapshot
2021.acl-long.447
A Closer Look at Few-Shot Crosslingual Transfer: The Choice of Shots Matters
https://aclanthology.org/2021.acl-long.447/
[ "Mengjie Zhao", "Yi Zhu", "Ehsan Shareghi", "Ivan Vulić", "Roi Reichart", "Anna Korhonen", "Hinrich Schütze" ]
Few-shot crosslingual transfer has been shown to outperform its zero-shot counterpart with pretrained encoders like multilingual BERT. Despite its growing popularity, little to no attention has been paid to standardizing and analyzing the design of few-shot experiments. In this work, we highlight a fundamental risk pos...
2021.acl-long.447
10.18653/v1/2021.acl-long.447
null
2012.15682
title_snapshot
2021.acl-long.448
Coreference Reasoning in Machine Reading Comprehension
https://aclanthology.org/2021.acl-long.448/
[ "Mingzhu Wu", "Nafise Sadat Moosavi", "Dan Roth", "Iryna Gurevych" ]
Coreference resolution is essential for natural language understanding and has been long studied in NLP. In recent years, as the format of Question Answering (QA) became a standard for machine reading comprehension (MRC), there have been data collection efforts, e.g., Dasigi et al. (2019), that attempt to evaluate the ...
2021.acl-long.448
10.18653/v1/2021.acl-long.448
null
2012.15573
title_snapshot
2021.acl-long.449
Adapting Unsupervised Syntactic Parsing Methodology for Discourse Dependency Parsing
https://aclanthology.org/2021.acl-long.449/
[ "Liwen Zhang", "Ge Wang", "Wenjuan Han", "Kewei Tu" ]
One of the main bottlenecks in developing discourse dependency parsers is the lack of annotated training data. A potential solution is to utilize abundant unlabeled data by using unsupervised techniques, but there is so far little research in unsupervised discourse dependency parsing. Fortunately, unsupervised syntacti...
2021.acl-long.449
10.18653/v1/2021.acl-long.449
null
null
null
2021.acl-long.450
A Conditional Splitting Framework for Efficient Constituency Parsing
https://aclanthology.org/2021.acl-long.450/
[ "Thanh-Tung Nguyen", "Xuan-Phi Nguyen", "Shafiq Joty", "Xiaoli Li" ]
We introduce a generic seq2seq parsing framework that casts constituency parsing problems (syntactic and discourse parsing) into a series of conditional splitting decisions. Our parsing model estimates the conditional probability distribution of possible splitting points in a given text span and supports efficient top-...
2021.acl-long.450
10.18653/v1/2021.acl-long.450
null
2106.15760
title_snapshot
2021.acl-long.451
A Unified Generative Framework for Various NER Subtasks
https://aclanthology.org/2021.acl-long.451/
[ "Hang Yan", "Tao Gui", "Junqi Dai", "Qipeng Guo", "Zheng Zhang", "Xipeng Qiu" ]
Named Entity Recognition (NER) is the task of identifying spans that represent entities in sentences. Whether the entity spans are nested or discontinuous, the NER task can be categorized into the flat NER, nested NER, and discontinuous NER subtasks. These subtasks have been mainly solved by the token-level sequence la...
2021.acl-long.451
10.18653/v1/2021.acl-long.451
null
2106.01223
title_snapshot
2021.acl-long.452
An In-depth Study on Internal Structure of Chinese Words
https://aclanthology.org/2021.acl-long.452/
[ "Chen Gong", "Saihao Huang", "Houquan Zhou", "Zhenghua Li", "Min Zhang", "Zhefeng Wang", "Baoxing Huai", "Nicholas Jing Yuan" ]
Unlike English letters, Chinese characters have rich and specific meanings. Usually, the meaning of a word can be derived from its constituent characters in some way. Several previous works on syntactic parsing propose to annotate shallow word-internal structures for better utilizing character-level information. This w...
2021.acl-long.452
10.18653/v1/2021.acl-long.452
null
2106.00334
title_snapshot
2021.acl-long.453
MulDA: A Multilingual Data Augmentation Framework for Low-Resource Cross-Lingual NER
https://aclanthology.org/2021.acl-long.453/
[ "Linlin Liu", "Bosheng Ding", "Lidong Bing", "Shafiq Joty", "Luo Si", "Chunyan Miao" ]
Named Entity Recognition (NER) for low-resource languages is a both practical and challenging research problem. This paper addresses zero-shot transfer for cross-lingual NER, especially when the amount of source-language training data is also limited. The paper first proposes a simple but effective labeled sequence tra...
2021.acl-long.453
10.18653/v1/2021.acl-long.453
null
null
null
2021.acl-long.454
Lexicon Enhanced Chinese Sequence Labeling Using BERT Adapter
https://aclanthology.org/2021.acl-long.454/
[ "Wei Liu", "Xiyan Fu", "Yue Zhang", "Wenming Xiao" ]
Lexicon information and pre-trained models, such as BERT, have been combined to explore Chinese sequence labeling tasks due to their respective strengths. However, existing methods solely fuse lexicon features via a shallow and random initialized sequence layer and do not integrate them into the bottom layers of BERT. ...
2021.acl-long.454
10.18653/v1/2021.acl-long.454
null
2105.07148
title_snapshot
2021.acl-long.455
Math Word Problem Solving with Explicit Numerical Values
https://aclanthology.org/2021.acl-long.455/
[ "Qinzhuo Wu", "Qi Zhang", "Zhongyu Wei", "Xuanjing Huang" ]
In recent years, math word problem solving has received considerable attention and achieved promising results, but previous methods rarely take numerical values into consideration. Most methods treat the numerical values in the problems as number symbols, and ignore the prominent role of the numerical values in solving...
2021.acl-long.455
10.18653/v1/2021.acl-long.455
null
null
null
2021.acl-long.456
Neural-Symbolic Solver for Math Word Problems with Auxiliary Tasks
https://aclanthology.org/2021.acl-long.456/
[ "Jinghui Qin", "Xiaodan Liang", "Yining Hong", "Jianheng Tang", "Liang Lin" ]
Previous math word problem solvers following the encoder-decoder paradigm fail to explicitly incorporate essential math symbolic constraints, leading to unexplainable and unreasonable predictions. Herein, we propose Neural-Symbolic Solver (NS-Solver) to explicitly and seamlessly incorporate different levels of symbolic...
2021.acl-long.456
10.18653/v1/2021.acl-long.456
null
2107.01431
title_snapshot
2021.acl-long.457
SMedBERT: A Knowledge-Enhanced Pre-trained Language Model with Structured Semantics for Medical Text Mining
https://aclanthology.org/2021.acl-long.457/
[ "Taolin Zhang", "Zerui Cai", "Chengyu Wang", "Minghui Qiu", "Bite Yang", "Xiaofeng He" ]
Recently, the performance of Pre-trained Language Models (PLMs) has been significantly improved by injecting knowledge facts to enhance their abilities of language understanding. For medical domains, the background knowledge sources are especially useful, due to the massive medical terms and their complicated relations...
2021.acl-long.457
10.18653/v1/2021.acl-long.457
null
2108.08983
title_snapshot
2021.acl-long.458
What is Your Article Based On? Inferring Fine-grained Provenance
https://aclanthology.org/2021.acl-long.458/
[ "Yi Zhang", "Zachary Ives", "Dan Roth" ]
When evaluating an article and the claims it makes, a critical reader must be able to assess where the information presented comes from, and whether the various claims are mutually consistent and support the conclusion. This motivates the study of claim provenance, which seeks to trace and explain the origins of claims...
2021.acl-long.458
10.18653/v1/2021.acl-long.458
null
null
null
2021.acl-long.459
Cross-modal Memory Networks for Radiology Report Generation
https://aclanthology.org/2021.acl-long.459/
[ "Zhihong Chen", "Yaling Shen", "Yan Song", "Xiang Wan" ]
Medical imaging plays a significant role in clinical practice of medical diagnosis, where the text reports of the images are essential in understanding them and facilitating later treatments. By generating the reports automatically, it is beneficial to help lighten the burden of radiologists and significantly promote c...
2021.acl-long.459
10.18653/v1/2021.acl-long.459
null
2204.13258
title_snapshot
2021.acl-long.460
Controversy and Conformity: from Generalized to Personalized Aggressiveness Detection
https://aclanthology.org/2021.acl-long.460/
[ "Kamil Kanclerz", "Alicja Figas", "Marcin Gruza", "Tomasz Kajdanowicz", "Jan Kocon", "Daria Puchalska", "Przemyslaw Kazienko" ]
There is content such as hate speech, offensive, toxic or aggressive documents, which are perceived differently by their consumers. They are commonly identified using classifiers solely based on textual content that generalize pre-agreed meanings of difficult problems. Such models provide the same results for each user...
2021.acl-long.460
10.18653/v1/2021.acl-long.460
null
null
null
2021.acl-long.461
Multi-perspective Coherent Reasoning for Helpfulness Prediction of Multimodal Reviews
https://aclanthology.org/2021.acl-long.461/
[ "Junhao Liu", "Zhen Hai", "Min Yang", "Lidong Bing" ]
As more and more product reviews are posted in both text and images, Multimodal Review Analysis (MRA) becomes an attractive research topic. Among the existing review analysis tasks, helpfulness prediction on review text has become predominant due to its importance for e-commerce platforms and online shops, i.e. helping...
2021.acl-long.461
10.18653/v1/2021.acl-long.461
null
null
null
2021.acl-long.462
Instantaneous Grammatical Error Correction with Shallow Aggressive Decoding
https://aclanthology.org/2021.acl-long.462/
[ "Xin Sun", "Tao Ge", "Furu Wei", "Houfeng Wang" ]
In this paper, we propose Shallow Aggressive Decoding (SAD) to improve the online inference efficiency of the Transformer for instantaneous Grammatical Error Correction (GEC). SAD optimizes the online inference efficiency for GEC by two innovations: 1) it aggressively decodes as many tokens as possible in parallel inst...
2021.acl-long.462
10.18653/v1/2021.acl-long.462
null
2106.04970
title_snapshot
2021.acl-long.463
Automatic ICD Coding via Interactive Shared Representation Networks with Self-distillation Mechanism
https://aclanthology.org/2021.acl-long.463/
[ "Tong Zhou", "Pengfei Cao", "Yubo Chen", "Kang Liu", "Jun Zhao", "Kun Niu", "Weifeng Chong", "Shengping Liu" ]
The ICD coding task aims at assigning codes of the International Classification of Diseases in clinical notes. Since manual coding is very laborious and prone to errors, many methods have been proposed for the automatic ICD coding task. However, existing works either ignore the long-tail of code frequency or the noisy ...
2021.acl-long.463
10.18653/v1/2021.acl-long.463
null
null
null
2021.acl-long.464
PHMOSpell: Phonological and Morphological Knowledge Guided Chinese Spelling Check
https://aclanthology.org/2021.acl-long.464/
[ "Li Huang", "Junjie Li", "Weiwei Jiang", "Zhiyu Zhang", "Minchuan Chen", "Shaojun Wang", "Jing Xiao" ]
Chinese Spelling Check (CSC) is a challenging task due to the complex characteristics of Chinese characters. Statistics reveal that most Chinese spelling errors belong to phonological or visual errors. However, previous methods rarely utilize phonological and morphological knowledge of Chinese characters or heavily rel...
2021.acl-long.464
10.18653/v1/2021.acl-long.464
null
null
null
2021.acl-long.465
Guiding the Growth: Difficulty-Controllable Question Generation through Step-by-Step Rewriting
https://aclanthology.org/2021.acl-long.465/
[ "Yi Cheng", "Siyao Li", "Bang Liu", "Ruihui Zhao", "Sujian Li", "Chenghua Lin", "Yefeng Zheng" ]
This paper explores the task of Difficulty-Controllable Question Generation (DCQG), which aims at generating questions with required difficulty levels. Previous research on this task mainly defines the difficulty of a question as whether it can be correctly answered by a Question Answering (QA) system, lacking interpre...
2021.acl-long.465
10.18653/v1/2021.acl-long.465
null
2105.11698
title_snapshot
2021.acl-long.466
Improving Encoder by Auxiliary Supervision Tasks for Table-to-Text Generation
https://aclanthology.org/2021.acl-long.466/
[ "Liang Li", "Can Ma", "Yinliang Yue", "Dayong Hu" ]
Table-to-text generation aims at automatically generating natural text to help people conveniently obtain salient information in tables. Although neural models for table-to-text have achieved remarkable progress, some problems are still overlooked. Previous methods cannot deduce the factual results from the entity’s (p...
2021.acl-long.466
10.18653/v1/2021.acl-long.466
null
null
null
2021.acl-long.467
POS-Constrained Parallel Decoding for Non-autoregressive Generation
https://aclanthology.org/2021.acl-long.467/
[ "Kexin Yang", "Wenqiang Lei", "Dayiheng Liu", "Weizhen Qi", "Jiancheng Lv" ]
The multimodality problem has become a major challenge of existing non-autoregressive generation (NAG) systems. A common solution often resorts to sequence-level knowledge distillation by rebuilding the training dataset through autoregressive generation (hereinafter known as “teacher AG”). The success of such methods m...
2021.acl-long.467
10.18653/v1/2021.acl-long.467
null
null
null
2021.acl-long.468
Bridging Subword Gaps in Pretrain-Finetune Paradigm for Natural Language Generation
https://aclanthology.org/2021.acl-long.468/
[ "Xin Liu", "Baosong Yang", "Dayiheng Liu", "Haibo Zhang", "Weihua Luo", "Min Zhang", "Haiying Zhang", "Jinsong Su" ]
A well-known limitation in pretrain-finetune paradigm lies in its inflexibility caused by the one-size-fits-all vocabulary. This potentially weakens the effect when applying pretrained models into natural language generation (NLG) tasks, especially for the subword distributions between upstream and downstream tasks wit...
2021.acl-long.468
10.18653/v1/2021.acl-long.468
null
2106.06125
title_snapshot
2021.acl-long.469
TGEA: An Error-Annotated Dataset and Benchmark Tasks for TextGeneration from Pretrained Language Models
https://aclanthology.org/2021.acl-long.469/
[ "Jie He", "Bo Peng", "Yi Liao", "Qun Liu", "Deyi Xiong" ]
In order to deeply understand the capability of pretrained language models in text generation and conduct a diagnostic evaluation, we propose TGEA, an error-annotated dataset with multiple benchmark tasks for text generation from pretrained language models (PLMs). We use carefully selected prompt words to guide GPT-2 t...
2021.acl-long.469
10.18653/v1/2021.acl-long.469
null
null
null
2021.acl-long.470
Long-Span Summarization via Local Attention and Content Selection
https://aclanthology.org/2021.acl-long.470/
[ "Potsawee Manakul", "Mark Gales" ]
Transformer-based models have achieved state-of-the-art results in a wide range of natural language processing (NLP) tasks including document summarization. Typically these systems are trained by fine-tuning a large pre-trained model to the target task. One issue with these transformer-based models is that they do not ...
2021.acl-long.470
10.18653/v1/2021.acl-long.470
null
2105.03801
title_snapshot
2021.acl-long.471
RepSum: Unsupervised Dialogue Summarization based on Replacement Strategy
https://aclanthology.org/2021.acl-long.471/
[ "Xiyan Fu", "Yating Zhang", "Tianyi Wang", "Xiaozhong Liu", "Changlong Sun", "Zhenglu Yang" ]
In the field of dialogue summarization, due to the lack of training data, it is often difficult for supervised summary generation methods to learn vital information from dialogue context with limited data. Several attempts on unsupervised summarization for text by leveraging semantic information solely or auto-encoder ...
2021.acl-long.471
10.18653/v1/2021.acl-long.471
null
null
null
2021.acl-long.472
BASS: Boosting Abstractive Summarization with Unified Semantic Graph
https://aclanthology.org/2021.acl-long.472/
[ "Wenhao Wu", "Wei Li", "Xinyan Xiao", "Jiachen Liu", "Ziqiang Cao", "Sujian Li", "Hua Wu", "Haifeng Wang" ]
Abstractive summarization for long-document or multi-document remains challenging for the Seq2Seq architecture, as Seq2Seq is not good at analyzing long-distance relations in text. In this paper, we present BASS, a novel framework for Boosting Abstractive Summarization based on a unified Semantic graph, which aggregate...
2021.acl-long.472
10.18653/v1/2021.acl-long.472
null
2105.12041
title_snapshot
2021.acl-long.473
Capturing Relations between Scientific Papers: An Abstractive Model for Related Work Section Generation
https://aclanthology.org/2021.acl-long.473/
[ "Xiuying Chen", "Hind Alamro", "Mingzhe Li", "Shen Gao", "Xiangliang Zhang", "Dongyan Zhao", "Rui Yan" ]
Given a set of related publications, related work section generation aims to provide researchers with an overview of the specific research area by summarizing these works and introducing them in a logical order. Most of existing related work generation models follow the inflexible extractive style, which directly extra...
2021.acl-long.473
10.18653/v1/2021.acl-long.473
null
null
null
2021.acl-long.474
Focus Attention: Promoting Faithfulness and Diversity in Summarization
https://aclanthology.org/2021.acl-long.474/
[ "Rahul Aralikatte", "Shashi Narayan", "Joshua Maynez", "Sascha Rothe", "Ryan McDonald" ]
Professional summaries are written with document-level information, such as the theme of the document, in mind. This is in contrast with most seq2seq decoders which simultaneously learn to focus on salient content, while deciding what to generate, at each decoding step. With the motivation to narrow this gap, we introd...
2021.acl-long.474
10.18653/v1/2021.acl-long.474
null
2105.11921
title_snapshot
2021.acl-long.475
Generating Query Focused Summaries from Query-Free Resources
https://aclanthology.org/2021.acl-long.475/
[ "Yumo Xu", "Mirella Lapata" ]
The availability of large-scale datasets has driven the development of neural models that create generic summaries from single or multiple documents. In this work we consider query focused summarization (QFS), a task for which training data in the form of queries, documents, and summaries is not readily available. We p...
2021.acl-long.475
10.18653/v1/2021.acl-long.475
null
2012.14774
title_snapshot
2021.acl-long.476
Robustifying Multi-hop QA through Pseudo-Evidentiality Training
https://aclanthology.org/2021.acl-long.476/
[ "Kyungjae Lee", "Seung-won Hwang", "Sang-eun Han", "Dohyeon Lee" ]
This paper studies the bias problem of multi-hop question answering models, of answering correctly without correct reasoning. One way to robustify these models is by supervising to not only answer right, but also with right reasoning chains. An existing direction is to annotate reasoning chains to train models, requiri...
2021.acl-long.476
10.18653/v1/2021.acl-long.476
null
2107.03242
title_snapshot
2021.acl-long.477
xMoCo: Cross Momentum Contrastive Learning for Open-Domain Question Answering
https://aclanthology.org/2021.acl-long.477/
[ "Nan Yang", "Furu Wei", "Binxing Jiao", "Daxing Jiang", "Linjun Yang" ]
Dense passage retrieval has been shown to be an effective approach for information retrieval tasks such as open domain question answering. Under this paradigm, a dual-encoder model is learned to encode questions and passages separately into vector representations, and all the passage vectors are then pre-computed and i...
2021.acl-long.477
10.18653/v1/2021.acl-long.477
null
null
null
2021.acl-long.478
Learn to Resolve Conversational Dependency: A Consistency Training Framework for Conversational Question Answering
https://aclanthology.org/2021.acl-long.478/
[ "Gangwoo Kim", "Hyunjae Kim", "Jungsoo Park", "Jaewoo Kang" ]
One of the main challenges in conversational question answering (CQA) is to resolve the conversational dependency, such as anaphora and ellipsis. However, existing approaches do not explicitly train QA models on how to resolve the dependency, and thus these models are limited in understanding human dialogues. In this p...
2021.acl-long.478
10.18653/v1/2021.acl-long.478
null
2106.11575
title_snapshot
2021.acl-long.479
PhotoChat: A Human-Human Dialogue Dataset With Photo Sharing Behavior For Joint Image-Text Modeling
https://aclanthology.org/2021.acl-long.479/
[ "Xiaoxue Zang", "Lijuan Liu", "Maria Wang", "Yang Song", "Hao Zhang", "Jindong Chen" ]
We present a new human-human dialogue dataset - PhotoChat, the first dataset that casts light on the photo sharing behavior in online messaging. PhotoChat contains 12k dialogues, each of which is paired with a user photo that is shared during the conversation. Based on this dataset, we propose two tasks to facilitate r...
2021.acl-long.479
10.18653/v1/2021.acl-long.479
null
2108.01453
title_snapshot
2021.acl-long.480
Good for Misconceived Reasons: An Empirical Revisiting on the Need for Visual Context in Multimodal Machine Translation
https://aclanthology.org/2021.acl-long.480/
[ "Zhiyong Wu", "Lingpeng Kong", "Wei Bi", "Xiang Li", "Ben Kao" ]
A neural multimodal machine translation (MMT) system is one that aims to perform better translation by extending conventional text-only translation models with multimodal information. Many recent studies report improvements when equipping their models with the multimodal module, despite the controversy of whether such ...
2021.acl-long.480
10.18653/v1/2021.acl-long.480
null
2105.14462
title_snapshot
2021.acl-long.481
Attend What You Need: Motion-Appearance Synergistic Networks for Video Question Answering
https://aclanthology.org/2021.acl-long.481/
[ "Ahjeong Seo", "Gi-Cheon Kang", "Joonhan Park", "Byoung-Tak Zhang" ]
Video Question Answering is a task which requires an AI agent to answer questions grounded in video. This task entails three key challenges: (1) understand the intention of various questions, (2) capturing various elements of the input video (e.g., object, action, causality), and (3) cross-modal grounding between langu...
2021.acl-long.481
10.18653/v1/2021.acl-long.481
null
2106.10446
title_snapshot
2021.acl-long.482
BERTifying the Hidden Markov Model for Multi-Source Weakly Supervised Named Entity Recognition
https://aclanthology.org/2021.acl-long.482/
[ "Yinghao Li", "Pranav Shetty", "Lucas Liu", "Chao Zhang", "Le Song" ]
We study the problem of learning a named entity recognition (NER) tagger using noisy labels from multiple weak supervision sources. Though cheap to obtain, the labels from weak supervision sources are often incomplete, inaccurate, and contradictory, making it difficult to learn an accurate NER model. To address this ch...
2021.acl-long.482
10.18653/v1/2021.acl-long.482
null
2105.12848
title_snapshot
2021.acl-long.483
CIL: Contrastive Instance Learning Framework for Distantly Supervised Relation Extraction
https://aclanthology.org/2021.acl-long.483/
[ "Tao Chen", "Haizhou Shi", "Siliang Tang", "Zhigang Chen", "Fei Wu", "Yueting Zhuang" ]
The journey of reducing noise from distant supervision (DS) generated training data has been started since the DS was first introduced into the relation extraction (RE) task. For the past decade, researchers apply the multi-instance learning (MIL) framework to find the most reliable feature from a bag of sentences. Alt...
2021.acl-long.483
10.18653/v1/2021.acl-long.483
null
2106.10855
title_snapshot
2021.acl-long.484
SENT: Sentence-level Distant Relation Extraction via Negative Training
https://aclanthology.org/2021.acl-long.484/
[ "Ruotian Ma", "Tao Gui", "Linyang Li", "Qi Zhang", "Xuanjing Huang", "Yaqian Zhou" ]
Distant supervision for relation extraction provides uniform bag labels for each sentence inside the bag, while accurate sentence labels are important for downstream applications that need the exact relation type. Directly using bag labels for sentence-level training will introduce much noise, thus severely degrading p...
2021.acl-long.484
10.18653/v1/2021.acl-long.484
null
2106.11566
title_snapshot
2021.acl-long.485
An End-to-End Progressive Multi-Task Learning Framework for Medical Named Entity Recognition and Normalization
https://aclanthology.org/2021.acl-long.485/
[ "Baohang Zhou", "Xiangrui Cai", "Ying Zhang", "Xiaojie Yuan" ]
Medical named entity recognition (NER) and normalization (NEN) are fundamental for constructing knowledge graphs and building QA systems. Existing implementations for medical NER and NEN are suffered from the error propagation between the two tasks. The mispredicted mentions from NER will directly influence the results...
2021.acl-long.485
10.18653/v1/2021.acl-long.485
null
null
null
2021.acl-long.486
PRGC: Potential Relation and Global Correspondence Based Joint Relational Triple Extraction
https://aclanthology.org/2021.acl-long.486/
[ "Hengyi Zheng", "Rui Wen", "Xi Chen", "Yifan Yang", "Yunyan Zhang", "Ziheng Zhang", "Ningyu Zhang", "Bin Qin", "Xu Ming", "Yefeng Zheng" ]
Joint extraction of entities and relations from unstructured texts is a crucial task in information extraction. Recent methods achieve considerable performance but still suffer from some inherent limitations, such as redundancy of relation prediction, poor generalization of span-based extraction and inefficiency. In th...
2021.acl-long.486
10.18653/v1/2021.acl-long.486
null
2106.09895
title_snapshot
2021.acl-long.487
Learning from Miscellaneous Other-Class Words for Few-shot Named Entity Recognition
https://aclanthology.org/2021.acl-long.487/
[ "Meihan Tong", "Shuai Wang", "Bin Xu", "Yixin Cao", "Minghui Liu", "Lei Hou", "Juanzi Li" ]
Few-shot Named Entity Recognition (NER) exploits only a handful of annotations to iden- tify and classify named entity mentions. Pro- totypical network shows superior performance on few-shot NER. However, existing prototyp- ical methods fail to differentiate rich seman- tics in other-class words, which will aggravate o...
2021.acl-long.487
10.18653/v1/2021.acl-long.487
null
2106.15167
title_snapshot
2021.acl-long.488
Joint Biomedical Entity and Relation Extraction with Knowledge-Enhanced Collective Inference
https://aclanthology.org/2021.acl-long.488/
[ "Tuan Lai", "Heng Ji", "ChengXiang Zhai", "Quan Hung Tran" ]
Compared to the general news domain, information extraction (IE) from biomedical text requires much broader domain knowledge. However, many previous IE methods do not utilize any external knowledge during inference. Due to the exponential growth of biomedical publications, models that do not go beyond their fixed set o...
2021.acl-long.488
10.18653/v1/2021.acl-long.488
null
2105.13456
title_snapshot
2021.acl-long.489
Fine-grained Information Extraction from Biomedical Literature based on Knowledge-enriched Abstract Meaning Representation
https://aclanthology.org/2021.acl-long.489/
[ "Zixuan Zhang", "Nikolaus Parulian", "Heng Ji", "Ahmed Elsayed", "Skatje Myers", "Martha Palmer" ]
Biomedical Information Extraction from scientific literature presents two unique and non-trivial challenges. First, compared with general natural language texts, sentences from scientific papers usually possess wider contexts between knowledge elements. Moreover, comprehending the fine-grained scientific entities and e...
2021.acl-long.489
10.18653/v1/2021.acl-long.489
null
null
null
2021.acl-long.490
Unleash GPT-2 Power for Event Detection
https://aclanthology.org/2021.acl-long.490/
[ "Amir Pouran Ben Veyseh", "Viet Lai", "Franck Dernoncourt", "Thien Huu Nguyen" ]
Event Detection (ED) aims to recognize mentions of events (i.e., event triggers) and their types in text. Recently, several ED datasets in various domains have been proposed. However, the major limitation of these resources is the lack of enough training data for individual event types which hinders the efficient train...
2021.acl-long.490
10.18653/v1/2021.acl-long.490
null
null
null
2021.acl-long.491
CLEVE: Contrastive Pre-training for Event Extraction
https://aclanthology.org/2021.acl-long.491/
[ "Ziqi Wang", "Xiaozhi Wang", "Xu Han", "Yankai Lin", "Lei Hou", "Zhiyuan Liu", "Peng Li", "Juanzi Li", "Jie Zhou" ]
Event extraction (EE) has considerably benefited from pre-trained language models (PLMs) by fine-tuning. However, existing pre-training methods have not involved modeling event characteristics, resulting in the developed EE models cannot take full advantage of large-scale unsupervised data. To this end, we propose CLEV...
2021.acl-long.491
10.18653/v1/2021.acl-long.491
null
2105.14485
title_snapshot
2021.acl-long.492
Document-level Event Extraction via Parallel Prediction Networks
https://aclanthology.org/2021.acl-long.492/
[ "Hang Yang", "Dianbo Sui", "Yubo Chen", "Kang Liu", "Jun Zhao", "Taifeng Wang" ]
Document-level event extraction (DEE) is indispensable when events are described throughout a document. We argue that sentence-level extractors are ill-suited to the DEE task where event arguments always scatter across sentences and multiple events may co-exist in a document. It is a challenging task because it require...
2021.acl-long.492
10.18653/v1/2021.acl-long.492
null
null
null
2021.acl-long.493
StructuralLM: Structural Pre-training for Form Understanding
https://aclanthology.org/2021.acl-long.493/
[ "Chenliang Li", "Bin Bi", "Ming Yan", "Wei Wang", "Songfang Huang", "Fei Huang", "Luo Si" ]
Large pre-trained language models achieve state-of-the-art results when fine-tuned on downstream NLP tasks. However, they almost exclusively focus on text-only representation, while neglecting cell-level layout information that is important for form image understanding. In this paper, we propose a new pre-training appr...
2021.acl-long.493
10.18653/v1/2021.acl-long.493
null
2105.11210
title_snapshot
2021.acl-long.494
Dual Graph Convolutional Networks for Aspect-based Sentiment Analysis
https://aclanthology.org/2021.acl-long.494/
[ "Ruifan Li", "Hao Chen", "Fangxiang Feng", "Zhanyu Ma", "Xiaojie Wang", "Eduard Hovy" ]
Aspect-based sentiment analysis is a fine-grained sentiment classification task. Recently, graph neural networks over dependency trees have been explored to explicitly model connections between aspects and opinion words. However, the improvement is limited due to the inaccuracy of the dependency parsing results and the...
2021.acl-long.494
10.18653/v1/2021.acl-long.494
null
null
null
2021.acl-long.495
Multi-Label Few-Shot Learning for Aspect Category Detection
https://aclanthology.org/2021.acl-long.495/
[ "Mengting Hu", "Shiwan Zhao", "Honglei Guo", "Chao Xue", "Hang Gao", "Tiegang Gao", "Renhong Cheng", "Zhong Su" ]
Aspect category detection (ACD) in sentiment analysis aims to identify the aspect categories mentioned in a sentence. In this paper, we formulate ACD in the few-shot learning scenario. However, existing few-shot learning approaches mainly focus on single-label predictions. These methods can not work well for the ACD ta...
2021.acl-long.495
10.18653/v1/2021.acl-long.495
null
2105.14174
title_snapshot
2021.acl-long.496
Argument Pair Extraction via Attention-guided Multi-Layer Multi-Cross Encoding
https://aclanthology.org/2021.acl-long.496/
[ "Liying Cheng", "Tianyu Wu", "Lidong Bing", "Luo Si" ]
Argument pair extraction (APE) is a research task for extracting arguments from two passages and identifying potential argument pairs. Prior research work treats this task as a sequence labeling problem and a binary classification problem on two passages that are directly concatenated together, which has a limitation o...
2021.acl-long.496
10.18653/v1/2021.acl-long.496
null
null
null
2021.acl-long.497
A Neural Transition-based Model for Argumentation Mining
https://aclanthology.org/2021.acl-long.497/
[ "Jianzhu Bao", "Chuang Fan", "Jipeng Wu", "Yixue Dang", "Jiachen Du", "Ruifeng Xu" ]
The goal of argumentation mining is to automatically extract argumentation structures from argumentative texts. Most existing methods determine argumentative relations by exhaustively enumerating all possible pairs of argument components, which suffer from low efficiency and class imbalance. Moreover, due to the comple...
2021.acl-long.497
10.18653/v1/2021.acl-long.497
null
null
null
2021.acl-long.498
Keep It Simple: Unsupervised Simplification of Multi-Paragraph Text
https://aclanthology.org/2021.acl-long.498/
[ "Philippe Laban", "Tobias Schnabel", "Paul Bennett", "Marti A. Hearst" ]
This work presents Keep it Simple (KiS), a new approach to unsupervised text simplification which learns to balance a reward across three properties: fluency, salience and simplicity. We train the model with a novel algorithm to optimize the reward (k-SCST), in which the model proposes several candidate simplifications...
2021.acl-long.498
10.18653/v1/2021.acl-long.498
null
2107.03444
title_snapshot
2021.acl-long.499
Long Text Generation by Modeling Sentence-Level and Discourse-Level Coherence
https://aclanthology.org/2021.acl-long.499/
[ "Jian Guan", "Xiaoxi Mao", "Changjie Fan", "Zitao Liu", "Wenbiao Ding", "Minlie Huang" ]
Generating long and coherent text is an important but challenging task, particularly for open-ended language generation tasks such as story generation. Despite the success in modeling intra-sentence coherence, existing generation models (e.g., BART) still struggle to maintain a coherent event sequence throughout the ge...
2021.acl-long.499
10.18653/v1/2021.acl-long.499
null
2105.08963
title_snapshot
2021.acl-long.500
OpenMEVA: A Benchmark for Evaluating Open-ended Story Generation Metrics
https://aclanthology.org/2021.acl-long.500/
[ "Jian Guan", "Zhexin Zhang", "Zhuoer Feng", "Zitao Liu", "Wenbiao Ding", "Xiaoxi Mao", "Changjie Fan", "Minlie Huang" ]
Automatic metrics are essential for developing natural language generation (NLG) models, particularly for open-ended language generation tasks such as story generation. However, existing automatic metrics are observed to correlate poorly with human evaluation. The lack of standardized benchmark datasets makes it diffic...
2021.acl-long.500
10.18653/v1/2021.acl-long.500
null
2105.08920
title_snapshot