paper_id
stringlengths
17
19
title
stringlengths
27
144
paper_url
stringlengths
43
45
authors
listlengths
1
37
abstract
large_stringlengths
396
1.69k
anthology_id
stringlengths
17
19
doi
stringlengths
29
31
award
stringclasses
0 values
arxiv_id
stringlengths
10
10
arxiv_id_source
stringclasses
2 values
2021.emnlp-main.301
FiD-Ex: Improving Sequence-to-Sequence Models for Extractive Rationale Generation
https://aclanthology.org/2021.emnlp-main.301/
[ "Kushal Lakhotia", "Bhargavi Paranjape", "Asish Ghoshal", "Scott Yih", "Yashar Mehdad", "Srini Iyer" ]
Natural language (NL) explanations of model predictions are gaining popularity as a means to understand and verify decisions made by large black-box pre-trained models, for tasks such as Question Answering (QA) and Fact Verification. Recently, pre-trained sequence to sequence (seq2seq) models have proven to be very eff...
2021.emnlp-main.301
10.18653/v1/2021.emnlp-main.301
null
2012.15482
title_snapshot
2021.emnlp-main.302
RockNER: A Simple Method to Create Adversarial Examples for Evaluating the Robustness of Named Entity Recognition Models
https://aclanthology.org/2021.emnlp-main.302/
[ "Bill Yuchen Lin", "Wenyang Gao", "Jun Yan", "Ryan Moreno", "Xiang Ren" ]
To audit the robustness of named entity recognition (NER) models, we propose RockNER, a simple yet effective method to create natural adversarial examples. Specifically, at the entity level, we replace target entities with other entities of the same semantic class in Wikidata; at the context level, we use pre-trained l...
2021.emnlp-main.302
10.18653/v1/2021.emnlp-main.302
null
2109.05620
title_snapshot
2021.emnlp-main.303
Diagnosing the First-Order Logical Reasoning Ability Through LogicNLI
https://aclanthology.org/2021.emnlp-main.303/
[ "Jidong Tian", "Yitian Li", "Wenqing Chen", "Liqiang Xiao", "Hao He", "Yaohui Jin" ]
Recently, language models (LMs) have achieved significant performance on many NLU tasks, which has spurred widespread interest for their possible applications in the scientific and social area. However, LMs have faced much criticism of whether they are truly capable of reasoning in NLU. In this work, we propose a diagn...
2021.emnlp-main.303
10.18653/v1/2021.emnlp-main.303
null
null
null
2021.emnlp-main.304
Constructing a Psychometric Testbed for Fair Natural Language Processing
https://aclanthology.org/2021.emnlp-main.304/
[ "Ahmed Abbasi", "David Dobolyi", "John P. Lalor", "Richard G. Netemeyer", "Kendall Smith", "Yi Yang" ]
Psychometric measures of ability, attitudes, perceptions, and beliefs are crucial for understanding user behavior in various contexts including health, security, e-commerce, and finance. Traditionally, psychometric dimensions have been measured and collected using survey-based methods. Inferring such constructs from us...
2021.emnlp-main.304
10.18653/v1/2021.emnlp-main.304
null
2007.12969
title_judge
2021.emnlp-main.305
COUGH: A Challenge Dataset and Models for COVID-19 FAQ Retrieval
https://aclanthology.org/2021.emnlp-main.305/
[ "Xinliang Frederick Zhang", "Heming Sun", "Xiang Yue", "Simon Lin", "Huan Sun" ]
We present a large, challenging dataset, COUGH, for COVID-19 FAQ retrieval. Similar to a standard FAQ dataset, COUGH consists of three parts: FAQ Bank, Query Bank and Relevance Set. The FAQ Bank contains ~16K FAQ items scraped from 55 credible websites (e.g., CDC and WHO). For evaluation, we introduce Query Bank and Re...
2021.emnlp-main.305
10.18653/v1/2021.emnlp-main.305
null
2010.12800
title_snapshot
2021.emnlp-main.306
Chinese WPLC: A Chinese Dataset for Evaluating Pretrained Language Models on Word Prediction Given Long-Range Context
https://aclanthology.org/2021.emnlp-main.306/
[ "Huibin Ge", "Chenxi Sun", "Deyi Xiong", "Qun Liu" ]
This paper presents a Chinese dataset for evaluating pretrained language models on Word Prediction given Long-term Context (Chinese WPLC). We propose both automatic and manual selection strategies tailored to Chinese to guarantee that target words in passages collected from over 69K novels can only be predicted with lo...
2021.emnlp-main.306
10.18653/v1/2021.emnlp-main.306
null
null
null
2021.emnlp-main.307
WinoLogic: A Zero-Shot Logic-based Diagnostic Dataset for Winograd Schema Challenge
https://aclanthology.org/2021.emnlp-main.307/
[ "Weinan He", "Canming Huang", "Yongmei Liu", "Xiaodan Zhu" ]
The recent success of neural language models (NLMs) on the Winograd Schema Challenge has called for further investigation of the commonsense reasoning ability of these models. Previous diagnostic datasets rely on crowd-sourcing which fails to provide coherent commonsense crucial for solving WSC problems. To better eval...
2021.emnlp-main.307
10.18653/v1/2021.emnlp-main.307
null
null
null
2021.emnlp-main.308
Pseudo Zero Pronoun Resolution Improves Zero Anaphora Resolution
https://aclanthology.org/2021.emnlp-main.308/
[ "Ryuto Konno", "Shun Kiyono", "Yuichiroh Matsubayashi", "Hiroki Ouchi", "Kentaro Inui" ]
Masked language models (MLMs) have contributed to drastic performance improvements with regard to zero anaphora resolution (ZAR). To further improve this approach, in this study, we made two proposals. The first is a new pretraining task that trains MLMs on anaphoric relations with explicit supervision, and the second ...
2021.emnlp-main.308
10.18653/v1/2021.emnlp-main.308
null
2104.07425
title_snapshot
2021.emnlp-main.309
Aligning Cross-lingual Sentence Representations with Dual Momentum Contrast
https://aclanthology.org/2021.emnlp-main.309/
[ "Liang Wang", "Wei Zhao", "Jingming Liu" ]
In this paper, we propose to align sentence representations from different languages into a unified embedding space, where semantic similarities (both cross-lingual and monolingual) can be computed with a simple dot product. Pre-trained language models are fine-tuned with the translation ranking task. Existing work (Fe...
2021.emnlp-main.309
10.18653/v1/2021.emnlp-main.309
null
2109.00253
title_snapshot
2021.emnlp-main.310
Total Recall: a Customized Continual Learning Method for Neural Semantic Parsers
https://aclanthology.org/2021.emnlp-main.310/
[ "Zhuang Li", "Lizhen Qu", "Gholamreza Haffari" ]
This paper investigates continual learning for semantic parsing. In this setting, a neural semantic parser learns tasks sequentially without accessing full training data from previous tasks. Direct application of the SOTA continual learning algorithms to this problem fails to achieve comparable performance with re-trai...
2021.emnlp-main.310
10.18653/v1/2021.emnlp-main.310
null
2109.05186
title_snapshot
2021.emnlp-main.311
Exophoric Pronoun Resolution in Dialogues with Topic Regularization
https://aclanthology.org/2021.emnlp-main.311/
[ "Xintong Yu", "Hongming Zhang", "Yangqiu Song", "Changshui Zhang", "Kun Xu", "Dong Yu" ]
Resolving pronouns to their referents has long been studied as a fundamental natural language understanding problem. Previous works on pronoun coreference resolution (PCR) mostly focus on resolving pronouns to mentions in text while ignoring the exophoric scenario. Exophoric pronouns are common in daily communications,...
2021.emnlp-main.311
10.18653/v1/2021.emnlp-main.311
null
2109.04787
title_snapshot
2021.emnlp-main.312
Context-Aware Interaction Network for Question Matching
https://aclanthology.org/2021.emnlp-main.312/
[ "Zhe Hu", "Zuohui Fu", "Yu Yin", "Gerard de Melo" ]
Impressive milestones have been achieved in text matching by adopting a cross-attention mechanism to capture pertinent semantic connections between two sentence representations. However, regular cross-attention focuses on word-level links between the two input sequences, neglecting the importance of contextual informat...
2021.emnlp-main.312
10.18653/v1/2021.emnlp-main.312
null
2104.08451
title_snapshot
2021.emnlp-main.313
TEMP: Taxonomy Expansion with Dynamic Margin Loss through Taxonomy-Paths
https://aclanthology.org/2021.emnlp-main.313/
[ "Zichen Liu", "Hongyuan Xu", "Yanlong Wen", "Ning Jiang", "HaiYing Wu", "Xiaojie Yuan" ]
As an essential form of knowledge representation, taxonomies are widely used in various downstream natural language processing tasks. However, with the continuously rising of new concepts, many existing taxonomies are unable to maintain coverage by manual expansion. In this paper, we propose TEMP, a self-supervised tax...
2021.emnlp-main.313
10.18653/v1/2021.emnlp-main.313
null
null
null
2021.emnlp-main.314
A Graph-Based Neural Model for End-to-End Frame Semantic Parsing
https://aclanthology.org/2021.emnlp-main.314/
[ "ZhiChao Lin", "Yueheng Sun", "Meishan Zhang" ]
Frame semantic parsing is a semantic analysis task based on FrameNet which has received great attention recently. The task usually involves three subtasks sequentially: (1) target identification, (2) frame classification and (3) semantic role labeling. The three subtasks are closely related while previous studies model...
2021.emnlp-main.314
10.18653/v1/2021.emnlp-main.314
null
2109.12319
title_snapshot
2021.emnlp-main.315
Virtual Data Augmentation: A Robust and General Framework for Fine-tuning Pre-trained Models
https://aclanthology.org/2021.emnlp-main.315/
[ "Kun Zhou", "Wayne Xin Zhao", "Sirui Wang", "Fuzheng Zhang", "Wei Wu", "Ji-Rong Wen" ]
Recent works have shown that powerful pre-trained language models (PLM) can be fooled by small perturbations or intentional attacks. To solve this issue, various data augmentation techniques are proposed to improve the robustness of PLMs. However, it is still challenging to augment semantically relevant examples with s...
2021.emnlp-main.315
10.18653/v1/2021.emnlp-main.315
null
2109.05793
title_snapshot
2021.emnlp-main.316
CATE: A Contrastive Pre-trained Model for Metaphor Detection with Semi-supervised Learning
https://aclanthology.org/2021.emnlp-main.316/
[ "Zhenxi Lin", "Qianli Ma", "Jiangyue Yan", "Jieyu Chen" ]
Metaphors are ubiquitous in natural language, and detecting them requires contextual reasoning about whether a semantic incongruence actually exists. Most existing work addresses this problem using pre-trained contextualized models. Despite their success, these models require a large amount of labeled data and are not ...
2021.emnlp-main.316
10.18653/v1/2021.emnlp-main.316
null
null
null
2021.emnlp-main.317
To be Closer: Learning to Link up Aspects with Opinions
https://aclanthology.org/2021.emnlp-main.317/
[ "Yuxiang Zhou", "Lejian Liao", "Yang Gao", "Zhanming Jie", "Wei Lu" ]
Dependency parse trees are helpful for discovering the opinion words in aspect-based sentiment analysis (ABSA) (CITATION). However, the trees obtained from off-the-shelf dependency parsers are static, and could be sub-optimal in ABSA. This is because the syntactic trees are not designed for capturing the interactions b...
2021.emnlp-main.317
10.18653/v1/2021.emnlp-main.317
null
2109.08382
title_snapshot
2021.emnlp-main.318
Seeking Common but Distinguishing Difference, A Joint Aspect-based Sentiment Analysis Model
https://aclanthology.org/2021.emnlp-main.318/
[ "Hongjiang Jing", "Zuchao Li", "Hai Zhao", "Shu Jiang" ]
Aspect-based sentiment analysis (ABSA) task consists of three typical subtasks: aspect term extraction, opinion term extraction, and sentiment polarity classification. These three subtasks are usually performed jointly to save resources and reduce the error propagation in the pipeline. However, most of the existing joi...
2021.emnlp-main.318
10.18653/v1/2021.emnlp-main.318
null
2111.09634
title_snapshot
2021.emnlp-main.319
Argument Pair Extraction with Mutual Guidance and Inter-sentence Relation Graph
https://aclanthology.org/2021.emnlp-main.319/
[ "Jianzhu Bao", "Bin Liang", "Jingyi Sun", "Yice Zhang", "Min Yang", "Ruifeng Xu" ]
Argument pair extraction (APE) aims to extract interactive argument pairs from two passages of a discussion. Previous work studied this task in the context of peer review and rebuttal, and decomposed it into a sequence labeling task and a sentence relation classification task. However, despite the promising performance...
2021.emnlp-main.319
10.18653/v1/2021.emnlp-main.319
null
null
null
2021.emnlp-main.320
Emotion Inference in Multi-Turn Conversations with Addressee-Aware Module and Ensemble Strategy
https://aclanthology.org/2021.emnlp-main.320/
[ "Dayu Li", "Xiaodan Zhu", "Yang Li", "Suge Wang", "Deyu Li", "Jian Liao", "Jianxing Zheng" ]
Emotion inference in multi-turn conversations aims to predict the participant’s emotion in the next upcoming turn without knowing the participant’s response yet, and is a necessary step for applications such as dialogue planning. However, it is a severe challenge to perceive and reason about the future feelings of part...
2021.emnlp-main.320
10.18653/v1/2021.emnlp-main.320
null
null
null
2021.emnlp-main.321
Improving Federated Learning for Aspect-based Sentiment Analysis via Topic Memories
https://aclanthology.org/2021.emnlp-main.321/
[ "Han Qin", "Guimin Chen", "Yuanhe Tian", "Yan Song" ]
Aspect-based sentiment analysis (ABSA) predicts the sentiment polarity towards a particular aspect term in a sentence, which is an important task in real-world applications. To perform ABSA, the trained model is required to have a good understanding of the contextual information, especially the particular patterns that...
2021.emnlp-main.321
10.18653/v1/2021.emnlp-main.321
null
null
null
2021.emnlp-main.322
Comparative Opinion Quintuple Extraction from Product Reviews
https://aclanthology.org/2021.emnlp-main.322/
[ "Ziheng Liu", "Rui Xia", "Jianfei Yu" ]
As an important task in opinion mining, comparative opinion mining aims to identify comparative sentences from product reviews, extract the comparative elements, and obtain the corresponding comparative opinion tuples. However, most previous studies simply regarded comparative tuple extraction as comparative element ex...
2021.emnlp-main.322
10.18653/v1/2021.emnlp-main.322
null
null
null
2021.emnlp-main.323
CTAL: Pre-training Cross-modal Transformer for Audio-and-Language Representations
https://aclanthology.org/2021.emnlp-main.323/
[ "Hang Li", "Wenbiao Ding", "Yu Kang", "Tianqiao Liu", "Zhongqin Wu", "Zitao Liu" ]
Existing audio-language task-specific predictive approaches focus on building complicated late-fusion mechanisms. However, these models are facing challenges of overfitting with limited labels and low model generalization abilities. In this paper, we present a Cross-modal Transformer for Audio-and-Language, i.e., CTAL,...
2021.emnlp-main.323
10.18653/v1/2021.emnlp-main.323
null
2109.00181
title_snapshot
2021.emnlp-main.324
Relation-aware Video Reading Comprehension for Temporal Language Grounding
https://aclanthology.org/2021.emnlp-main.324/
[ "Jialin Gao", "Xin Sun", "Mengmeng Xu", "Xi Zhou", "Bernard Ghanem" ]
Temporal language grounding in videos aims to localize the temporal span relevant to the given query sentence. Previous methods treat it either as a boundary regression task or a span extraction task. This paper will formulate temporal language grounding into video reading comprehension and propose a Relation-aware Net...
2021.emnlp-main.324
10.18653/v1/2021.emnlp-main.324
null
2110.05717
title_snapshot
2021.emnlp-main.325
Mutual-Learning Improves End-to-End Speech Translation
https://aclanthology.org/2021.emnlp-main.325/
[ "Jiawei Zhao", "Wei Luo", "Boxing Chen", "Andrew Gilman" ]
A currently popular research area in end-to-end speech translation is the use of knowledge distillation from a machine translation (MT) task to improve the speech translation (ST) task. However, such scenario obviously only allows one way transfer, which is limited by the performance of the teacher model. Therefore, We...
2021.emnlp-main.325
10.18653/v1/2021.emnlp-main.325
null
null
null
2021.emnlp-main.326
Vision Guided Generative Pre-trained Language Models for Multimodal Abstractive Summarization
https://aclanthology.org/2021.emnlp-main.326/
[ "Tiezheng Yu", "Wenliang Dai", "Zihan Liu", "Pascale Fung" ]
Multimodal abstractive summarization (MAS) models that summarize videos (vision modality) and their corresponding transcripts (text modality) are able to extract the essential information from massive multimodal data on the Internet. Recently, large-scale generative pre-trained language models (GPLMs) have been shown t...
2021.emnlp-main.326
10.18653/v1/2021.emnlp-main.326
null
2109.02401
title_snapshot
2021.emnlp-main.327
Natural Language Video Localization with Learnable Moment Proposals
https://aclanthology.org/2021.emnlp-main.327/
[ "Shaoning Xiao", "Long Chen", "Jian Shao", "Yueting Zhuang", "Jun Xiao" ]
Given an untrimmed video and a natural language query, Natural Language Video Localization (NLVL) aims to identify the video moment described by query. To address this task, existing methods can be roughly grouped into two groups: 1) propose-and-rank models first define a set of hand-designed moment candidates and then...
2021.emnlp-main.327
10.18653/v1/2021.emnlp-main.327
null
2109.10678
title_snapshot
2021.emnlp-main.328
Language-Aligned Waypoint (LAW) Supervision for Vision-and-Language Navigation in Continuous Environments
https://aclanthology.org/2021.emnlp-main.328/
[ "Sonia Raychaudhuri", "Saim Wani", "Shivansh Patel", "Unnat Jain", "Angel Chang" ]
In the Vision-and-Language Navigation (VLN) task an embodied agent navigates a 3D environment, following natural language instructions. A challenge in this task is how to handle ‘off the path’ scenarios where an agent veers from a reference path. Prior work supervises the agent with actions based on the shortest path f...
2021.emnlp-main.328
10.18653/v1/2021.emnlp-main.328
null
2109.15207
title_snapshot
2021.emnlp-main.329
How to leverage the multimodal EHR data for better medical prediction?
https://aclanthology.org/2021.emnlp-main.329/
[ "Bo Yang", "Lijun Wu" ]
Healthcare is becoming a more and more important research topic recently. With the growing data in the healthcare domain, it offers a great opportunity for deep learning to improve the quality of service and reduce costs. However, the complexity of electronic health records (EHR) data is a challenge for the application...
2021.emnlp-main.329
10.18653/v1/2021.emnlp-main.329
null
2110.15763
title_judge
2021.emnlp-main.330
Considering Nested Tree Structure in Sentence Extractive Summarization with Pre-trained Transformer
https://aclanthology.org/2021.emnlp-main.330/
[ "Jingun Kwon", "Naoki Kobayashi", "Hidetaka Kamigaito", "Manabu Okumura" ]
Sentence extractive summarization shortens a document by selecting sentences for a summary while preserving its important contents. However, constructing a coherent and informative summary is difficult using a pre-trained BERT-based encoder since it is not explicitly trained for representing the information of sentence...
2021.emnlp-main.330
10.18653/v1/2021.emnlp-main.330
null
null
null
2021.emnlp-main.331
Frame Semantic-Enhanced Sentence Modeling for Sentence-level Extractive Text Summarization
https://aclanthology.org/2021.emnlp-main.331/
[ "Yong Guan", "Shaoru Guo", "Ru Li", "Xiaoli Li", "Hongye Tan" ]
Sentence-level extractive text summarization aims to select important sentences from a given document. However, it is very challenging to model the importance of sentences. In this paper, we propose a novel Frame Semantic-Enhanced Sentence Modeling for Extractive Summarization, which leverages Frame semantics to model ...
2021.emnlp-main.331
10.18653/v1/2021.emnlp-main.331
null
null
null
2021.emnlp-main.332
CAST: Enhancing Code Summarization with Hierarchical Splitting and Reconstruction of Abstract Syntax Trees
https://aclanthology.org/2021.emnlp-main.332/
[ "Ensheng Shi", "Yanlin Wang", "Lun Du", "Hongyu Zhang", "Shi Han", "Dongmei Zhang", "Hongbin Sun" ]
Code summarization aims to generate concise natural language descriptions of source code, which can help improve program comprehension and maintenance. Recent studies show that syntactic and structural information extracted from abstract syntax trees (ASTs) is conducive to summary generation. However, existing approach...
2021.emnlp-main.332
10.18653/v1/2021.emnlp-main.332
null
2108.12987
title_snapshot
2021.emnlp-main.333
SgSum:Transforming Multi-document Summarization into Sub-graph Selection
https://aclanthology.org/2021.emnlp-main.333/
[ "Moye Chen", "Wei Li", "Jiachen Liu", "Xinyan Xiao", "Hua Wu", "Haifeng Wang" ]
Most of existing extractive multi-document summarization (MDS) methods score each sentence individually and extract salient sentences one by one to compose a summary, which have two main drawbacks: (1) neglecting both the intra and cross-document relations between sentences; (2) neglecting the coherence and conciseness...
2021.emnlp-main.333
10.18653/v1/2021.emnlp-main.333
null
2110.12645
title_snapshot
2021.emnlp-main.334
Event Graph based Sentence Fusion
https://aclanthology.org/2021.emnlp-main.334/
[ "Ruifeng Yuan", "Zili Wang", "Wenjie Li" ]
Sentence fusion is a conditional generation task that merges several related sentences into a coherent one, which can be deemed as a summary sentence. The importance of sentence fusion has long been recognized by communities in natural language generation, especially in text summarization. It remains challenging for a ...
2021.emnlp-main.334
10.18653/v1/2021.emnlp-main.334
null
null
null
2021.emnlp-main.335
Transformer-based Lexically Constrained Headline Generation
https://aclanthology.org/2021.emnlp-main.335/
[ "Kosuke Yamada", "Yuta Hitomi", "Hideaki Tamori", "Ryohei Sasano", "Naoaki Okazaki", "Kentaro Inui", "Koichi Takeda" ]
This paper explores a variant of automatic headline generation methods, where a generated headline is required to include a given phrase such as a company or a product name. Previous methods using Transformer-based models generate a headline including a given phrase by providing the encoder with additional information ...
2021.emnlp-main.335
10.18653/v1/2021.emnlp-main.335
null
2109.07080
title_snapshot
2021.emnlp-main.336
Learn to Copy from the Copying History: Correlational Copy Network for Abstractive Summarization
https://aclanthology.org/2021.emnlp-main.336/
[ "Haoran Li", "Song Xu", "Peng Yuan", "Yujia Wang", "Youzheng Wu", "Xiaodong He", "Bowen Zhou" ]
The copying mechanism has had considerable success in abstractive summarization, facilitating models to directly copy words from the input text to the output summary. Existing works mostly employ encoder-decoder attention, which applies copying at each time step independently of the former ones. However, this may somet...
2021.emnlp-main.336
10.18653/v1/2021.emnlp-main.336
null
null
null
2021.emnlp-main.337
Gradient-Based Adversarial Factual Consistency Evaluation for Abstractive Summarization
https://aclanthology.org/2021.emnlp-main.337/
[ "Zhiyuan Zeng", "Jiaze Chen", "Weiran Xu", "Lei Li" ]
Neural abstractive summarization systems have gained significant progress in recent years. However, abstractive summarization often produce inconsisitent statements or false facts. How to automatically generate highly abstract yet factually correct summaries? In this paper, we proposed an efficient weak-supervised adve...
2021.emnlp-main.337
10.18653/v1/2021.emnlp-main.337
null
null
null
2021.emnlp-main.338
Word Reordering for Zero-shot Cross-lingual Structured Prediction
https://aclanthology.org/2021.emnlp-main.338/
[ "Tao Ji", "Yong Jiang", "Tao Wang", "Zhongqiang Huang", "Fei Huang", "Yuanbin Wu", "Xiaoling Wang" ]
Adapting word order from one language to another is a key problem in cross-lingual structured prediction. Current sentence encoders (e.g., RNN, Transformer with position embeddings) are usually word order sensitive. Even with uniform word form representations (MUSE, mBERT), word order discrepancies may hurt the adaptat...
2021.emnlp-main.338
10.18653/v1/2021.emnlp-main.338
null
null
null
2021.emnlp-main.339
A Unified Encoding of Structures in Transition Systems
https://aclanthology.org/2021.emnlp-main.339/
[ "Tao Ji", "Yong Jiang", "Tao Wang", "Zhongqiang Huang", "Fei Huang", "Yuanbin Wu", "Xiaoling Wang" ]
Transition systems usually contain various dynamic structures (e.g., stacks, buffers). An ideal transition-based model should encode these structures completely and efficiently. Previous works relying on templates or neural network structures either only encode partial structure information or suffer from computation e...
2021.emnlp-main.339
10.18653/v1/2021.emnlp-main.339
null
null
null
2021.emnlp-main.340
Improving Unsupervised Question Answering via Summarization-Informed Question Generation
https://aclanthology.org/2021.emnlp-main.340/
[ "Chenyang Lyu", "Lifeng Shang", "Yvette Graham", "Jennifer Foster", "Xin Jiang", "Qun Liu" ]
Question Generation (QG) is the task of generating a plausible question for a given <passage, answer> pair. Template-based QG uses linguistically-informed heuristics to transform declarative sentences into interrogatives, whereas supervised QG uses existing Question Answering (QA) datasets to train a system to generate...
2021.emnlp-main.340
10.18653/v1/2021.emnlp-main.340
null
2109.07954
title_snapshot
2021.emnlp-main.341
TransferNet: An Effective and Transparent Framework for Multi-hop Question Answering over Relation Graph
https://aclanthology.org/2021.emnlp-main.341/
[ "Jiaxin Shi", "Shulin Cao", "Lei Hou", "Juanzi Li", "Hanwang Zhang" ]
Multi-hop Question Answering (QA) is a challenging task because it requires precise reasoning with entity relations at every step towards the answer. The relations can be represented in terms of labels in knowledge graph (e.g., spouse) or text in text corpus (e.g., they have been married for 26 years). Existing models ...
2021.emnlp-main.341
10.18653/v1/2021.emnlp-main.341
null
2104.07302
title_snapshot
2021.emnlp-main.342
Topic Transferable Table Question Answering
https://aclanthology.org/2021.emnlp-main.342/
[ "Saneem Chemmengath", "Vishwajeet Kumar", "Samarth Bharadwaj", "Jaydeep Sen", "Mustafa Canim", "Soumen Chakrabarti", "Alfio Gliozzo", "Karthik Sankaranarayanan" ]
Weakly-supervised table question-answering (TableQA) models have achieved state-of-art performance by using pre-trained BERT transformer to jointly encoding a question and a table to produce structured query for the question. However, in practical settings TableQA systems are deployed over table corpora having topic an...
2021.emnlp-main.342
10.18653/v1/2021.emnlp-main.342
null
2109.07377
title_snapshot
2021.emnlp-main.343
WebSRC: A Dataset for Web-Based Structural Reading Comprehension
https://aclanthology.org/2021.emnlp-main.343/
[ "Xingyu Chen", "Zihan Zhao", "Lu Chen", "JiaBao Ji", "Danyang Zhang", "Ao Luo", "Yuxuan Xiong", "Kai Yu" ]
Web search is an essential way for humans to obtain information, but it’s still a great challenge for machines to understand the contents of web pages. In this paper, we introduce the task of web-based structural reading comprehension. Given a web page and a question about it, the task is to find an answer from the web...
2021.emnlp-main.343
10.18653/v1/2021.emnlp-main.343
null
2101.09465
title_snapshot
2021.emnlp-main.344
Cryptonite: A Cryptic Crossword Benchmark for Extreme Ambiguity in Language
https://aclanthology.org/2021.emnlp-main.344/
[ "Avia Efrat", "Uri Shaham", "Dan Kilman", "Omer Levy" ]
Current NLP datasets targeting ambiguity can be solved by a native speaker with relative ease. We present Cryptonite, a large-scale dataset based on cryptic crosswords, which is both linguistically complex and naturally sourced. Each example in Cryptonite is a cryptic clue, a short phrase or sentence with a misleading ...
2021.emnlp-main.344
10.18653/v1/2021.emnlp-main.344
null
2103.01242
title_snapshot
2021.emnlp-main.345
End-to-End Entity Resolution and Question Answering Using Differentiable Knowledge Graphs
https://aclanthology.org/2021.emnlp-main.345/
[ "Amir Saffari", "Armin Oliya", "Priyanka Sen", "Tom Ayoola" ]
Recently, end-to-end (E2E) trained models for question answering over knowledge graphs (KGQA) have delivered promising results using only a weakly supervised dataset. However, these models are trained and evaluated in a setting where hand-annotated question entities are supplied to the model, leaving the important and ...
2021.emnlp-main.345
10.18653/v1/2021.emnlp-main.345
null
2109.05817
title_snapshot
2021.emnlp-main.346
Improving Query Graph Generation for Complex Question Answering over Knowledge Base
https://aclanthology.org/2021.emnlp-main.346/
[ "Kechen Qin", "Cheng Li", "Virgil Pavlu", "Javed Aslam" ]
Most of the existing Knowledge-based Question Answering (KBQA) methods first learn to map the given question to a query graph, and then convert the graph to an executable query to find the answer. The query graph is typically expanded progressively from the topic entity based on a sequence prediction model. In this pap...
2021.emnlp-main.346
10.18653/v1/2021.emnlp-main.346
null
null
null
2021.emnlp-main.347
DiscoDVT: Generating Long Text with Discourse-Aware Discrete Variational Transformer
https://aclanthology.org/2021.emnlp-main.347/
[ "Haozhe Ji", "Minlie Huang" ]
Despite the recent advances in applying pre-trained language models to generate high-quality texts, generating long passages that maintain long-range coherence is yet challenging for these models. In this paper, we propose DiscoDVT, a discourse-aware discrete variational Transformer to tackle the incoherence issue. Dis...
2021.emnlp-main.347
10.18653/v1/2021.emnlp-main.347
null
2110.05999
title_snapshot
2021.emnlp-main.348
Mathematical Word Problem Generation from Commonsense Knowledge Graph and Equations
https://aclanthology.org/2021.emnlp-main.348/
[ "Tianqiao Liu", "Qiang Fang", "Wenbiao Ding", "Hang Li", "Zhongqin Wu", "Zitao Liu" ]
There is an increasing interest in the use of mathematical word problem (MWP) generation in educational assessment. Different from standard natural question generation, MWP generation needs to maintain the underlying mathematical operations between quantities and variables, while at the same time ensuring the relevance...
2021.emnlp-main.348
10.18653/v1/2021.emnlp-main.348
null
2010.06196
title_snapshot
2021.emnlp-main.349
Generic resources are what you need: Style transfer tasks without task-specific parallel training data
https://aclanthology.org/2021.emnlp-main.349/
[ "Huiyuan Lai", "Antonio Toral", "Malvina Nissim" ]
Style transfer aims to rewrite a source text in a different target style while preserving its content. We propose a novel approach to this task that leverages generic resources, and without using any task-specific parallel (source–target) data outperforms existing unsupervised approaches on the two most popular style t...
2021.emnlp-main.349
10.18653/v1/2021.emnlp-main.349
null
2109.04543
title_snapshot
2021.emnlp-main.350
Revisiting Pivot-Based Paraphrase Generation: Language Is Not the Only Optional Pivot
https://aclanthology.org/2021.emnlp-main.350/
[ "Yitao Cai", "Yue Cao", "Xiaojun Wan" ]
Paraphrases refer to texts that convey the same meaning with different expression forms. Pivot-based methods, also known as the round-trip translation, have shown promising results in generating high-quality paraphrases. However, existing pivot-based methods all rely on language as the pivot, where large-scale, high-qu...
2021.emnlp-main.350
10.18653/v1/2021.emnlp-main.350
null
null
null
2021.emnlp-main.351
Structural Adapters in Pretrained Language Models for AMR-to-Text Generation
https://aclanthology.org/2021.emnlp-main.351/
[ "Leonardo F. R. Ribeiro", "Yue Zhang", "Iryna Gurevych" ]
Pretrained language models (PLM) have recently advanced graph-to-text generation, where the input graph is linearized into a sequence and fed into the PLM to obtain its representation. However, efficiently encoding the graph structure in PLMs is challenging because such models were pretrained on natural language, and m...
2021.emnlp-main.351
10.18653/v1/2021.emnlp-main.351
null
2103.09120
title_snapshot
2021.emnlp-main.352
Data-to-text Generation by Splicing Together Nearest Neighbors
https://aclanthology.org/2021.emnlp-main.352/
[ "Sam Wiseman", "Arturs Backurs", "Karl Stratos" ]
We propose to tackle data-to-text generation tasks by directly splicing together retrieved segments of text from “neighbor” source-target pairs. Unlike recent work that conditions on retrieved neighbors but generates text token-by-token, left-to-right, we learn a policy that directly manipulates segments of neighbor te...
2021.emnlp-main.352
10.18653/v1/2021.emnlp-main.352
null
2101.08248
title_snapshot
2021.emnlp-main.353
Contextualize Knowledge Bases with Transformer for End-to-end Task-Oriented Dialogue Systems
https://aclanthology.org/2021.emnlp-main.353/
[ "Yanjie Gou", "Yinjie Lei", "Lingqiao Liu", "Yong Dai", "Chunxu Shen" ]
Incorporating knowledge bases (KB) into end-to-end task-oriented dialogue systems is challenging, since it requires to properly represent the entity of KB, which is associated with its KB context and dialogue context. The existing works represent the entity with only perceiving a part of its KB context, which can lead ...
2021.emnlp-main.353
10.18653/v1/2021.emnlp-main.353
null
2010.05740
title_snapshot
2021.emnlp-main.354
Efficient Dialogue Complementary Policy Learning via Deep Q-network Policy and Episodic Memory Policy
https://aclanthology.org/2021.emnlp-main.354/
[ "Yangyang Zhao", "Zhenyu Wang", "Changxi Zhu", "Shihan Wang" ]
Deep reinforcement learning has shown great potential in training dialogue policies. However, its favorable performance comes at the cost of many rounds of interaction. Most of the existing dialogue policy methods rely on a single learning system, while the human brain has two specialized learning and memory systems, s...
2021.emnlp-main.354
10.18653/v1/2021.emnlp-main.354
null
null
null
2021.emnlp-main.355
CRFR: Improving Conversational Recommender Systems via Flexible Fragments Reasoning on Knowledge Graphs
https://aclanthology.org/2021.emnlp-main.355/
[ "Jinfeng Zhou", "Bo Wang", "Ruifang He", "Yuexian Hou" ]
Although paths of user interests shift in knowledge graphs (KGs) can benefit conversational recommender systems (CRS), explicit reasoning on KGs has not been well considered in CRS, due to the complex of high-order and incomplete paths. We propose CRFR, which effectively does explicit multi-hop reasoning on KGs with a ...
2021.emnlp-main.355
10.18653/v1/2021.emnlp-main.355
null
null
null
2021.emnlp-main.356
DuRecDial 2.0: A Bilingual Parallel Corpus for Conversational Recommendation
https://aclanthology.org/2021.emnlp-main.356/
[ "Zeming Liu", "Haifeng Wang", "Zheng-Yu Niu", "Hua Wu", "Wanxiang Che" ]
In this paper, we provide a bilingual parallel human-to-human recommendation dialog dataset (DuRecDial 2.0) to enable researchers to explore a challenging task of multilingual and cross-lingual conversational recommendation. The difference between DuRecDial 2.0 and existing conversational recommendation datasets is tha...
2021.emnlp-main.356
10.18653/v1/2021.emnlp-main.356
null
2109.08877
title_snapshot
2021.emnlp-main.357
End-to-End Learning of Flowchart Grounded Task-Oriented Dialogs
https://aclanthology.org/2021.emnlp-main.357/
[ "Dinesh Raghu", "Shantanu Agarwal", "Sachindra Joshi", "Mausam" ]
We propose a novel problem within end-to-end learning of task oriented dialogs (TOD), in which the dialog system mimics a troubleshooting agent who helps a user by diagnosing their problem (e.g., car not starting). Such dialogs are grounded in domain-specific flowcharts, which the agent is supposed to follow during the...
2021.emnlp-main.357
10.18653/v1/2021.emnlp-main.357
null
2109.07263
title_snapshot
2021.emnlp-main.358
Dimensional Emotion Detection from Categorical Emotion
https://aclanthology.org/2021.emnlp-main.358/
[ "Sungjoon Park", "Jiseon Kim", "Seonghyeon Ye", "Jaeyeol Jeon", "Hee Young Park", "Alice Oh" ]
We present a model to predict fine-grained emotions along the continuous dimensions of valence, arousal, and dominance (VAD) with a corpus with categorical emotion annotations. Our model is trained by minimizing the EMD (Earth Mover’s Distance) loss between the predicted VAD score distribution and the categorical emoti...
2021.emnlp-main.358
10.18653/v1/2021.emnlp-main.358
null
1911.02499
title_snapshot
2021.emnlp-main.359
Not All Negatives are Equal: Label-Aware Contrastive Loss for Fine-grained Text Classification
https://aclanthology.org/2021.emnlp-main.359/
[ "Varsha Suresh", "Desmond C. Ong" ]
Fine-grained classification involves dealing with datasets with larger number of classes with subtle differences between them. Guiding the model to focus on differentiating dimensions between these commonly confusable classes is key to improving performance on fine-grained tasks. In this work, we analyse the contrastiv...
2021.emnlp-main.359
10.18653/v1/2021.emnlp-main.359
null
2109.05427
title_snapshot
2021.emnlp-main.360
Joint Multi-modal Aspect-Sentiment Analysis with Auxiliary Cross-modal Relation Detection
https://aclanthology.org/2021.emnlp-main.360/
[ "Xincheng Ju", "Dong Zhang", "Rong Xiao", "Junhui Li", "Shoushan Li", "Min Zhang", "Guodong Zhou" ]
Aspect terms extraction (ATE) and aspect sentiment classification (ASC) are two fundamental and fine-grained sub-tasks in aspect-level sentiment analysis (ALSA). In the textual analysis, joint extracting both aspect terms and sentiment polarities has been drawn much attention due to the better applications than individ...
2021.emnlp-main.360
10.18653/v1/2021.emnlp-main.360
null
null
null
2021.emnlp-main.361
Solving Aspect Category Sentiment Analysis as a Text Generation Task
https://aclanthology.org/2021.emnlp-main.361/
[ "Jian Liu", "Zhiyang Teng", "Leyang Cui", "Hanmeng Liu", "Yue Zhang" ]
Aspect category sentiment analysis has attracted increasing research attention. The dominant methods make use of pre-trained language models by learning effective aspect category-specific representations, and adding specific output layers to its pre-trained representation. We consider a more direct way of making use of...
2021.emnlp-main.361
10.18653/v1/2021.emnlp-main.361
null
2110.07310
title_snapshot
2021.emnlp-main.362
Semantics-Preserved Data Augmentation for Aspect-Based Sentiment Analysis
https://aclanthology.org/2021.emnlp-main.362/
[ "Ting-Wei Hsu", "Chung-Chi Chen", "Hen-Hsen Huang", "Hsin-Hsi Chen" ]
Both the issues of data deficiencies and semantic consistency are important for data augmentation. Most of previous methods address the first issue, but ignore the second one. In the cases of aspect-based sentiment analysis, violation of the above issues may change the aspect and sentiment polarity. In this paper, we p...
2021.emnlp-main.362
10.18653/v1/2021.emnlp-main.362
null
null
null
2021.emnlp-main.363
The Effect of Round-Trip Translation on Fairness in Sentiment Analysis
https://aclanthology.org/2021.emnlp-main.363/
[ "Jonathan Gabel Christiansen", "Mathias Gammelgaard", "Anders Søgaard" ]
Sentiment analysis systems have been shown to exhibit sensitivity to protected attributes. Round-trip translation, on the other hand, has been shown to normalize text. We explore the impact of round-trip translation on the demographic parity of sentiment classifiers and show how round-trip translation consistently impr...
2021.emnlp-main.363
10.18653/v1/2021.emnlp-main.363
null
null
null
2021.emnlp-main.364
CHoRaL: Collecting Humor Reaction Labels from Millions of Social Media Users
https://aclanthology.org/2021.emnlp-main.364/
[ "Zixiaofan Yang", "Shayan Hooshmand", "Julia Hirschberg" ]
Humor detection has gained attention in recent years due to the desire to understand user-generated content with figurative language. However, substantial individual and cultural differences in humor perception make it very difficult to collect a large-scale humor dataset with reliable humor labels. We propose CHoRaL, ...
2021.emnlp-main.364
10.18653/v1/2021.emnlp-main.364
null
null
null
2021.emnlp-main.365
CSDS: A Fine-Grained Chinese Dataset for Customer Service Dialogue Summarization
https://aclanthology.org/2021.emnlp-main.365/
[ "Haitao Lin", "Liqun Ma", "Junnan Zhu", "Lu Xiang", "Yu Zhou", "Jiajun Zhang", "Chengqing Zong" ]
Dialogue summarization has drawn much attention recently. Especially in the customer service domain, agents could use dialogue summaries to help boost their works by quickly knowing customer’s issues and service progress. These applications require summaries to contain the perspective of a single speaker and have a cle...
2021.emnlp-main.365
10.18653/v1/2021.emnlp-main.365
null
2108.13139
title_snapshot
2021.emnlp-main.366
CodRED: A Cross-Document Relation Extraction Dataset for Acquiring Knowledge in the Wild
https://aclanthology.org/2021.emnlp-main.366/
[ "Yuan Yao", "Jiaju Du", "Yankai Lin", "Peng Li", "Zhiyuan Liu", "Jie Zhou", "Maosong Sun" ]
Existing relation extraction (RE) methods typically focus on extracting relational facts between entity pairs within single sentences or documents. However, a large quantity of relational facts in knowledge bases can only be inferred across documents in practice. In this work, we present the problem of cross-document R...
2021.emnlp-main.366
10.18653/v1/2021.emnlp-main.366
null
null
null
2021.emnlp-main.367
Building and Evaluating Open-Domain Dialogue Corpora with Clarifying Questions
https://aclanthology.org/2021.emnlp-main.367/
[ "Mohammad Aliannejadi", "Julia Kiseleva", "Aleksandr Chuklin", "Jeff Dalton", "Mikhail Burtsev" ]
Enabling open-domain dialogue systems to ask clarifying questions when appropriate is an important direction for improving the quality of the system response. Namely, for cases when a user request is not specific enough for a conversation system to provide an answer right away, it is desirable to ask a clarifying quest...
2021.emnlp-main.367
10.18653/v1/2021.emnlp-main.367
null
2109.05794
title_snapshot
2021.emnlp-main.368
We Need to Talk About train-dev-test Splits
https://aclanthology.org/2021.emnlp-main.368/
[ "Rob van der Goot" ]
Standard train-dev-test splits used to benchmark multiple models against each other are ubiquitously used in Natural Language Processing (NLP). In this setup, the train data is used for training the model, the development set for evaluating different versions of the proposed model(s) during development, and the test se...
2021.emnlp-main.368
10.18653/v1/2021.emnlp-main.368
null
null
null
2021.emnlp-main.369
PhoMT: A High-Quality and Large-Scale Benchmark Dataset for Vietnamese-English Machine Translation
https://aclanthology.org/2021.emnlp-main.369/
[ "Long Doan", "Linh The Nguyen", "Nguyen Luong Tran", "Thai Hoang", "Dat Quoc Nguyen" ]
We introduce a high-quality and large-scale Vietnamese-English parallel dataset of 3.02M sentence pairs, which is 2.9M pairs larger than the benchmark Vietnamese-English machine translation corpus IWSLT15. We conduct experiments comparing strong neural baselines and well-known automatic translation engines on our datas...
2021.emnlp-main.369
10.18653/v1/2021.emnlp-main.369
null
2110.12199
title_snapshot
2021.emnlp-main.370
Lying Through One’s Teeth: A Study on Verbal Leakage Cues
https://aclanthology.org/2021.emnlp-main.370/
[ "Min-Hsuan Yeh", "Lun-Wei Ku" ]
Although many studies use the LIWC lexicon to show the existence of verbal leakage cues in lie detection datasets, none mention how verbal leakage cues are influenced by means of data collection, or the impact thereof on the performance of models. In this paper, we study verbal leakage cues to understand the effect of ...
2021.emnlp-main.370
10.18653/v1/2021.emnlp-main.370
null
null
null
2021.emnlp-main.371
Multi-granularity Textual Adversarial Attack with Behavior Cloning
https://aclanthology.org/2021.emnlp-main.371/
[ "Yangyi Chen", "Jin Su", "Wei Wei" ]
Recently, the textual adversarial attack models become increasingly popular due to their successful in estimating the robustness of NLP models. However, existing works have obvious deficiencies. (1)They usually consider only a single granularity of modification strategies (e.g. word-level or sentence-level), which is i...
2021.emnlp-main.371
10.18653/v1/2021.emnlp-main.371
null
2109.04367
title_snapshot
2021.emnlp-main.372
All Bark and No Bite: Rogue Dimensions in Transformer Language Models Obscure Representational Quality
https://aclanthology.org/2021.emnlp-main.372/
[ "William Timkey", "Marten van Schijndel" ]
Similarity measures are a vital tool for understanding how language models represent and process language. Standard representational similarity measures such as cosine similarity and Euclidean distance have been successfully used in static word embedding models to understand how words cluster in semantic space. Recentl...
2021.emnlp-main.372
10.18653/v1/2021.emnlp-main.372
null
2109.04404
title_snapshot
2021.emnlp-main.373
Incorporating Residual and Normalization Layers into Analysis of Masked Language Models
https://aclanthology.org/2021.emnlp-main.373/
[ "Goro Kobayashi", "Tatsuki Kuribayashi", "Sho Yokoi", "Kentaro Inui" ]
Transformer architecture has become ubiquitous in the natural language processing field. To interpret the Transformer-based models, their attention patterns have been extensively analyzed. However, the Transformer architecture is not only composed of the multi-head attention; other components can also contribute to Tra...
2021.emnlp-main.373
10.18653/v1/2021.emnlp-main.373
null
2109.07152
title_snapshot
2021.emnlp-main.374
Mind the Style of Text! Adversarial and Backdoor Attacks Based on Text Style Transfer
https://aclanthology.org/2021.emnlp-main.374/
[ "Fanchao Qi", "Yangyi Chen", "Xurui Zhang", "Mukai Li", "Zhiyuan Liu", "Maosong Sun" ]
Adversarial attacks and backdoor attacks are two common security threats that hang over deep learning. Both of them harness task-irrelevant features of data in their implementation. Text style is a feature that is naturally irrelevant to most NLP tasks, and thus suitable for adversarial and backdoor attacks. In this pa...
2021.emnlp-main.374
10.18653/v1/2021.emnlp-main.374
null
2110.07139
title_snapshot
2021.emnlp-main.375
Sociolectal Analysis of Pretrained Language Models
https://aclanthology.org/2021.emnlp-main.375/
[ "Sheng Zhang", "Xin Zhang", "Weiming Zhang", "Anders Søgaard" ]
Using data from English cloze tests, in which subjects also self-reported their gender, age, education, and race, we examine performance differences of pretrained language models across demographic groups, defined by these (protected) attributes. We demonstrate wide performance gaps across demographic groups and show t...
2021.emnlp-main.375
10.18653/v1/2021.emnlp-main.375
null
null
null
2021.emnlp-main.376
Examining Cross-lingual Contextual Embeddings with Orthogonal Structural Probes
https://aclanthology.org/2021.emnlp-main.376/
[ "Tomasz Limisiewicz", "David Mareček" ]
State-of-the-art contextual embeddings are obtained from large language models available only for a few languages. For others, we need to learn representations using a multilingual model. There is an ongoing debate on whether multilingual embeddings can be aligned in a space shared across many languages. The novel Orth...
2021.emnlp-main.376
10.18653/v1/2021.emnlp-main.376
null
2109.04921
title_snapshot
2021.emnlp-main.377
Are Transformers a Modern Version of ELIZA? Observations on French Object Verb Agreement
https://aclanthology.org/2021.emnlp-main.377/
[ "Bingzhi Li", "Guillaume Wisniewski", "Benoit Crabbé" ]
Many recent works have demonstrated that unsupervised sentence representations of neural networks encode syntactic information by observing that neural language models are able to predict the agreement between a verb and its subject. We take a critical look at this line of research by showing that it is possible to ach...
2021.emnlp-main.377
10.18653/v1/2021.emnlp-main.377
null
2109.10133
title_snapshot
2021.emnlp-main.378
Fine-grained Entity Typing via Label Reasoning
https://aclanthology.org/2021.emnlp-main.378/
[ "Qing Liu", "Hongyu Lin", "Xinyan Xiao", "Xianpei Han", "Le Sun", "Hua Wu" ]
Conventional entity typing approaches are based on independent classification paradigms, which make them difficult to recognize inter-dependent, long-tailed and fine-grained entity types. In this paper, we argue that the implicitly entailed extrinsic and intrinsic dependencies between labels can provide critical knowle...
2021.emnlp-main.378
10.18653/v1/2021.emnlp-main.378
null
2109.05744
title_snapshot
2021.emnlp-main.379
Enhanced Language Representation with Label Knowledge for Span Extraction
https://aclanthology.org/2021.emnlp-main.379/
[ "Pan Yang", "Xin Cong", "Zhenyu Sun", "Xingwu Liu" ]
Span extraction, aiming to extract text spans (such as words or phrases) from plain text, is a fundamental process in Information Extraction. Recent works introduce the label knowledge to enhance the text representation by formalizing the span extraction task into a question answering problem (QA Formalization), which ...
2021.emnlp-main.379
10.18653/v1/2021.emnlp-main.379
null
2111.00884
title_snapshot
2021.emnlp-main.380
PRIDE: Predicting Relationships in Conversations
https://aclanthology.org/2021.emnlp-main.380/
[ "Anna Tigunova", "Paramita Mirza", "Andrew Yates", "Gerhard Weikum" ]
Automatically extracting interpersonal relationships of conversation interlocutors can enrich personal knowledge bases to enhance personalized search, recommenders and chatbots. To infer speakers’ relationships from dialogues we propose PRIDE, a neural multi-label classifier, based on BERT and Transformer for creating ...
2021.emnlp-main.380
10.18653/v1/2021.emnlp-main.380
null
null
null
2021.emnlp-main.381
Extracting Fine-Grained Knowledge Graphs of Scientific Claims: Dataset and Transformer-Based Results
https://aclanthology.org/2021.emnlp-main.381/
[ "Ian Magnusson", "Scott Friedman" ]
Recent transformer-based approaches demonstrate promising results on relational scientific information extraction. Existing datasets focus on high-level description of how research is carried out. Instead we focus on the subtleties of how experimental associations are presented by building SciClaim, a dataset of scient...
2021.emnlp-main.381
10.18653/v1/2021.emnlp-main.381
null
2109.10453
title_snapshot
2021.emnlp-main.382
Sequential Cross-Document Coreference Resolution
https://aclanthology.org/2021.emnlp-main.382/
[ "Emily Allaway", "Shuai Wang", "Miguel Ballesteros" ]
Relating entities and events in text is a key component of natural language understanding. Cross-document coreference resolution, in particular, is important for the growing interest in multi-document analysis tasks. In this work we propose a new model that extends the efficient sequential prediction paradigm for coref...
2021.emnlp-main.382
10.18653/v1/2021.emnlp-main.382
null
2104.08413
title_snapshot
2021.emnlp-main.383
Mixture-of-Partitions: Infusing Large Biomedical Knowledge Graphs into BERT
https://aclanthology.org/2021.emnlp-main.383/
[ "Zaiqiao Meng", "Fangyu Liu", "Thomas Clark", "Ehsan Shareghi", "Nigel Collier" ]
Infusing factual knowledge into pre-trained models is fundamental for many knowledge-intensive tasks. In this paper, we proposed Mixture-of-Partitions (MoP), an infusion approach that can handle a very large knowledge graph (KG) by partitioning it into smaller sub-graphs and infusing their specific knowledge into vario...
2021.emnlp-main.383
10.18653/v1/2021.emnlp-main.383
null
2109.04810
title_snapshot
2021.emnlp-main.384
Filling the Gaps in Ancient Akkadian Texts: A Masked Language Modelling Approach
https://aclanthology.org/2021.emnlp-main.384/
[ "Koren Lazar", "Benny Saret", "Asaf Yehudai", "Wayne Horowitz", "Nathan Wasserman", "Gabriel Stanovsky" ]
We present models which complete missing text given transliterations of ancient Mesopotamian documents, originally written on cuneiform clay tablets (2500 BCE - 100 CE). Due to the tablets’ deterioration, scholars often rely on contextual cues to manually fill in missing parts in the text in a subjective and time-consu...
2021.emnlp-main.384
10.18653/v1/2021.emnlp-main.384
null
2109.04513
title_snapshot
2021.emnlp-main.385
AVocaDo: Strategy for Adapting Vocabulary to Downstream Domain
https://aclanthology.org/2021.emnlp-main.385/
[ "Jimin Hong", "TaeHee Kim", "Hyesu Lim", "Jaegul Choo" ]
During the fine-tuning phase of transfer learning, the pretrained vocabulary remains unchanged, while model parameters are updated. The vocabulary generated based on the pretrained data is suboptimal for downstream data when domain discrepancy exists. We propose to consider the vocabulary as an optimizable parameter, a...
2021.emnlp-main.385
10.18653/v1/2021.emnlp-main.385
null
2110.13434
title_snapshot
2021.emnlp-main.386
Can We Improve Model Robustness through Secondary Attribute Counterfactuals?
https://aclanthology.org/2021.emnlp-main.386/
[ "Ananth Balashankar", "Xuezhi Wang", "Ben Packer", "Nithum Thain", "Ed Chi", "Alex Beutel" ]
Developing robust NLP models that perform well on many, even small, slices of data is a significant but important challenge, with implications from fairness to general reliability. To this end, recent research has explored how models rely on spurious correlations, and how counterfactual data augmentation (CDA) can miti...
2021.emnlp-main.386
10.18653/v1/2021.emnlp-main.386
null
null
null
2021.emnlp-main.387
Long-Range Modeling of Source Code Files with eWASH: Extended Window Access by Syntax Hierarchy
https://aclanthology.org/2021.emnlp-main.387/
[ "Colin Clement", "Shuai Lu", "Xiaoyu Liu", "Michele Tufano", "Dawn Drain", "Nan Duan", "Neel Sundaresan", "Alexey Svyatkovskiy" ]
Statistical language modeling and translation with transformers have found many successful applications in program understanding and generation tasks, setting high benchmarks for tools in modern software development environments. The finite context window of these neural models means, however, that they will be unable ...
2021.emnlp-main.387
10.18653/v1/2021.emnlp-main.387
null
2109.08780
title_snapshot
2021.emnlp-main.388
Can Language Models be Biomedical Knowledge Bases?
https://aclanthology.org/2021.emnlp-main.388/
[ "Mujeen Sung", "Jinhyuk Lee", "Sean Yi", "Minji Jeon", "Sungdong Kim", "Jaewoo Kang" ]
Pre-trained language models (LMs) have become ubiquitous in solving various natural language processing (NLP) tasks. There has been increasing interest in what knowledge these LMs contain and how we can extract that knowledge, treating LMs as knowledge bases (KBs). While there has been much work on probing LMs in the g...
2021.emnlp-main.388
10.18653/v1/2021.emnlp-main.388
null
2109.07154
title_snapshot
2021.emnlp-main.389
LayoutReader: Pre-training of Text and Layout for Reading Order Detection
https://aclanthology.org/2021.emnlp-main.389/
[ "Zilong Wang", "Yiheng Xu", "Lei Cui", "Jingbo Shang", "Furu Wei" ]
Reading order detection is the cornerstone to understanding visually-rich documents (e.g., receipts and forms). Unfortunately, no existing work took advantage of advanced deep learning models because it is too laborious to annotate a large enough dataset. We observe that the reading order of WORD documents is embedded ...
2021.emnlp-main.389
10.18653/v1/2021.emnlp-main.389
null
2108.11591
title_snapshot
2021.emnlp-main.390
Region under Discussion for visual dialog
https://aclanthology.org/2021.emnlp-main.390/
[ "Mauricio Mazuecos", "Franco M. Luque", "Jorge Sánchez", "Hernán Maina", "Thomas Vadora", "Luciana Benotti" ]
Visual Dialog is assumed to require the dialog history to generate correct responses during a dialog. However, it is not clear from previous work how dialog history is needed for visual dialog. In this paper we define what it means for a visual question to require dialog history and we release a subset of the Guesswhat...
2021.emnlp-main.390
10.18653/v1/2021.emnlp-main.390
null
null
null
2021.emnlp-main.391
Learning grounded word meaning representations on similarity graphs
https://aclanthology.org/2021.emnlp-main.391/
[ "Mariella Dimiccoli", "Herwig Wendt", "Pau Batlle Franch" ]
This paper introduces a novel approach to learn visually grounded meaning representations of words as low-dimensional node embeddings on an underlying graph hierarchy. The lower level of the hierarchy models modality-specific word representations, conditioned to another modality, through dedicated but communicating gra...
2021.emnlp-main.391
10.18653/v1/2021.emnlp-main.391
null
2109.03084
title_snapshot
2021.emnlp-main.392
WhyAct: Identifying Action Reasons in Lifestyle Vlogs
https://aclanthology.org/2021.emnlp-main.392/
[ "Oana Ignat", "Santiago Castro", "Hanwen Miao", "Weiji Li", "Rada Mihalcea" ]
We aim to automatically identify human action reasons in online videos. We focus on the widespread genre of lifestyle vlogs, in which people perform actions while verbally describing them. We introduce and make publicly available the WhyAct dataset, consisting of 1,077 visual actions manually annotated with their reaso...
2021.emnlp-main.392
10.18653/v1/2021.emnlp-main.392
null
2109.02747
title_snapshot
2021.emnlp-main.393
Genre as Weak Supervision for Cross-lingual Dependency Parsing
https://aclanthology.org/2021.emnlp-main.393/
[ "Max Müller-Eberstein", "Rob van der Goot", "Barbara Plank" ]
Recent work has shown that monolingual masked language models learn to represent data-driven notions of language variation which can be used for domain-targeted training data selection. Dataset genre labels are already frequently available, yet remain largely unexplored in cross-lingual setups. We harness this genre me...
2021.emnlp-main.393
10.18653/v1/2021.emnlp-main.393
null
2109.04733
title_snapshot
2021.emnlp-main.394
On the Relation between Syntactic Divergence and Zero-Shot Performance
https://aclanthology.org/2021.emnlp-main.394/
[ "Ofir Arviv", "Dmitry Nikolaev", "Taelin Karidi", "Omri Abend" ]
We explore the link between the extent to which syntactic relations are preserved in translation and the ease of correctly constructing a parse tree in a zero-shot setting. While previous work suggests such a relation, it tends to focus on the macro level and not on the level of individual edges—a gap we aim to address...
2021.emnlp-main.394
10.18653/v1/2021.emnlp-main.394
null
2110.04644
title_snapshot
2021.emnlp-main.395
Improved Latent Tree Induction with Distant Supervision via Span Constraints
https://aclanthology.org/2021.emnlp-main.395/
[ "Zhiyang Xu", "Andrew Drozdov", "Jay Yoon Lee", "Tim O’Gorman", "Subendhu Rongali", "Dylan Finkbeiner", "Shilpa Suresh", "Mohit Iyyer", "Andrew McCallum" ]
For over thirty years, researchers have developed and analyzed methods for latent tree induction as an approach for unsupervised syntactic parsing. Nonetheless, modern systems still do not perform well enough compared to their supervised counterparts to have any practical use as structural annotation of text. In this w...
2021.emnlp-main.395
10.18653/v1/2021.emnlp-main.395
null
2109.05112
title_snapshot
2021.emnlp-main.396
Aligning Multidimensional Worldviews and Discovering Ideological Differences
https://aclanthology.org/2021.emnlp-main.396/
[ "Jeremiah Milbauer", "Adarsh Mathew", "James Evans" ]
The Internet is home to thousands of communities, each with their own unique worldview and associated ideological differences. With new communities constantly emerging and serving as ideological birthplaces, battlegrounds, and bunkers, it is critical to develop a framework for understanding worldviews and ideological d...
2021.emnlp-main.396
10.18653/v1/2021.emnlp-main.396
null
null
null
2021.emnlp-main.397
Just Say No: Analyzing the Stance of Neural Dialogue Generation in Offensive Contexts
https://aclanthology.org/2021.emnlp-main.397/
[ "Ashutosh Baheti", "Maarten Sap", "Alan Ritter", "Mark Riedl" ]
Dialogue models trained on human conversations inadvertently learn to generate toxic responses. In addition to producing explicitly offensive utterances, these models can also implicitly insult a group or individual by aligning themselves with an offensive statement. To better understand the dynamics of contextually of...
2021.emnlp-main.397
10.18653/v1/2021.emnlp-main.397
null
2108.11830
title_snapshot
2021.emnlp-main.398
Multi-Modal Open-Domain Dialogue
https://aclanthology.org/2021.emnlp-main.398/
[ "Kurt Shuster", "Eric Michael Smith", "Da Ju", "Jason Weston" ]
Recent work in open-domain conversational agents has demonstrated that significant improvements in humanness and user preference can be achieved via massive scaling in both pre-training data and model size (Adiwardana et al., 2020; Roller et al., 2020). However, if we want to build agents with human-like abilities, we ...
2021.emnlp-main.398
10.18653/v1/2021.emnlp-main.398
null
2010.01082
title_snapshot
2021.emnlp-main.399
A Label-Aware BERT Attention Network for Zero-Shot Multi-Intent Detection in Spoken Language Understanding
https://aclanthology.org/2021.emnlp-main.399/
[ "Ting-Wei Wu", "Ruolin Su", "Biing Juang" ]
With the early success of query-answer assistants such as Alexa and Siri, research attempts to expand system capabilities of handling service automation are now abundant. However, preliminary systems have quickly found the inadequacy in relying on simple classification techniques to effectively accomplish the automatio...
2021.emnlp-main.399
10.18653/v1/2021.emnlp-main.399
null
null
null
2021.emnlp-main.400
Zero-Shot Dialogue Disentanglement by Self-Supervised Entangled Response Selection
https://aclanthology.org/2021.emnlp-main.400/
[ "Ta-Chung Chi", "Alexander Rudnicky" ]
Dialogue disentanglement aims to group utterances in a long and multi-participant dialogue into threads. This is useful for discourse analysis and downstream applications such as dialogue response selection, where it can be the first step to construct a clean context/response set. Unfortunately, labeling all reply-to l...
2021.emnlp-main.400
10.18653/v1/2021.emnlp-main.400
null
2110.12646
title_snapshot