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.501
DYPLOC: Dynamic Planning of Content Using Mixed Language Models for Text Generation
https://aclanthology.org/2021.acl-long.501/
[ "Xinyu Hua", "Ashwin Sreevatsa", "Lu Wang" ]
We study the task of long-form opinion text generation, which faces at least two distinct challenges. First, existing neural generation models fall short of coherence, thus requiring efficient content planning. Second, diverse types of information are needed to guide the generator to cover both subjective and objective...
2021.acl-long.501
10.18653/v1/2021.acl-long.501
null
2106.00791
title_snapshot
2021.acl-long.502
Controllable Open-ended Question Generation with A New Question Type Ontology
https://aclanthology.org/2021.acl-long.502/
[ "Shuyang Cao", "Lu Wang" ]
We investigate the less-explored task of generating open-ended questions that are typically answered by multiple sentences. We first define a new question type ontology which differentiates the nuanced nature of questions better than widely used question words. A new dataset with 4,959 questions is labeled based on the...
2021.acl-long.502
10.18653/v1/2021.acl-long.502
null
2107.00152
title_snapshot
2021.acl-long.503
BERTGen: Multi-task Generation through BERT
https://aclanthology.org/2021.acl-long.503/
[ "Faidon Mitzalis", "Ozan Caglayan", "Pranava Madhyastha", "Lucia Specia" ]
We present BERTGen, a novel, generative, decoder-only model which extends BERT by fusing multimodal and multilingual pre-trained models VL-BERT and M-BERT, respectively. BERTGen is auto-regressively trained for language generation tasks, namely image captioning, machine translation and multimodal machine translation, u...
2021.acl-long.503
10.18653/v1/2021.acl-long.503
null
2106.03484
title_snapshot
2021.acl-long.504
Selective Knowledge Distillation for Neural Machine Translation
https://aclanthology.org/2021.acl-long.504/
[ "Fusheng Wang", "Jianhao Yan", "Fandong Meng", "Jie Zhou" ]
Neural Machine Translation (NMT) models achieve state-of-the-art performance on many translation benchmarks. As an active research field in NMT, knowledge distillation is widely applied to enhance the model’s performance by transferring teacher model’s knowledge on each training sample. However, previous work rarely di...
2021.acl-long.504
10.18653/v1/2021.acl-long.504
null
2105.12967
title_snapshot
2021.acl-long.505
Measuring and Increasing Context Usage in Context-Aware Machine Translation
https://aclanthology.org/2021.acl-long.505/
[ "Patrick Fernandes", "Kayo Yin", "Graham Neubig", "André F. T. Martins" ]
Recent work in neural machine translation has demonstrated both the necessity and feasibility of using inter-sentential context, context from sentences other than those currently being translated. However, while many current methods present model architectures that theoretically can use this extra context, it is often ...
2021.acl-long.505
10.18653/v1/2021.acl-long.505
null
2105.03482
title_snapshot
2021.acl-long.506
Beyond Offline Mapping: Learning Cross-lingual Word Embeddings through Context Anchoring
https://aclanthology.org/2021.acl-long.506/
[ "Aitor Ormazabal", "Mikel Artetxe", "Aitor Soroa", "Gorka Labaka", "Eneko Agirre" ]
Recent research on cross-lingual word embeddings has been dominated by unsupervised mapping approaches that align monolingual embeddings. Such methods critically rely on those embeddings having a similar structure, but it was recently shown that the separate training in different languages causes departures from this a...
2021.acl-long.506
10.18653/v1/2021.acl-long.506
null
2012.15715
title_snapshot
2021.acl-long.507
CCMatrix: Mining Billions of High-Quality Parallel Sentences on the Web
https://aclanthology.org/2021.acl-long.507/
[ "Holger Schwenk", "Guillaume Wenzek", "Sergey Edunov", "Edouard Grave", "Armand Joulin", "Angela Fan" ]
We show that margin-based bitext mining in a multilingual sentence space can be successfully scaled to operate on monolingual corpora of billions of sentences. We use 32 snapshots of a curated common crawl corpus (Wenzel et al, 2019) totaling 71 billion unique sentences. Using one unified approach for 90 languages, we ...
2021.acl-long.507
10.18653/v1/2021.acl-long.507
null
1911.04944
title_snapshot
2021.acl-long.508
Length-Adaptive Transformer: Train Once with Length Drop, Use Anytime with Search
https://aclanthology.org/2021.acl-long.508/
[ "Gyuwan Kim", "Kyunghyun Cho" ]
Despite transformers’ impressive accuracy, their computational cost is often prohibitive to use with limited computational resources. Most previous approaches to improve inference efficiency require a separate model for each possible computational budget. In this paper, we extend PoWER-BERT (Goyal et al., 2020) and pro...
2021.acl-long.508
10.18653/v1/2021.acl-long.508
null
2010.07003
title_snapshot
2021.acl-long.509
GhostBERT: Generate More Features with Cheap Operations for BERT
https://aclanthology.org/2021.acl-long.509/
[ "Zhiqi Huang", "Lu Hou", "Lifeng Shang", "Xin Jiang", "Xiao Chen", "Qun Liu" ]
Transformer-based pre-trained language models like BERT, though powerful in many tasks, are expensive in both memory and computation, due to their large number of parameters. Previous works show that some parameters in these models can be pruned away without severe accuracy drop. However, these redundant features contr...
2021.acl-long.509
10.18653/v1/2021.acl-long.509
null
null
null
2021.acl-long.510
Super Tickets in Pre-Trained Language Models: From Model Compression to Improving Generalization
https://aclanthology.org/2021.acl-long.510/
[ "Chen Liang", "Simiao Zuo", "Minshuo Chen", "Haoming Jiang", "Xiaodong Liu", "Pengcheng He", "Tuo Zhao", "Weizhu Chen" ]
The Lottery Ticket Hypothesis suggests that an over-parametrized network consists of ”lottery tickets”, and training a certain collection of them (i.e., a subnetwork) can match the performance of the full model. In this paper, we study such a collection of tickets, which is referred to as ”winning tickets”, in extremel...
2021.acl-long.510
10.18653/v1/2021.acl-long.510
null
2105.12002
title_snapshot
2021.acl-long.511
A Novel Estimator of Mutual Information for Learning to Disentangle Textual Representations
https://aclanthology.org/2021.acl-long.511/
[ "Pierre Colombo", "Pablo Piantanida", "Chloé Clavel" ]
Learning disentangled representations of textual data is essential for many natural language tasks such as fair classification, style transfer and sentence generation, among others. The existent dominant approaches in the context of text data either rely on training an adversary (discriminator) that aims at making attr...
2021.acl-long.511
10.18653/v1/2021.acl-long.511
null
2105.02685
title_snapshot
2021.acl-long.512
Determinantal Beam Search
https://aclanthology.org/2021.acl-long.512/
[ "Clara Meister", "Martina Forster", "Ryan Cotterell" ]
Beam search is a go-to strategy for decoding neural sequence models. The algorithm can naturally be viewed as a subset optimization problem, albeit one where the corresponding set function does not reflect interactions between candidates. Empirically, this leads to sets often exhibiting high overlap, e.g., strings may ...
2021.acl-long.512
10.18653/v1/2021.acl-long.512
null
2106.07400
title_snapshot
2021.acl-long.513
Multi-hop Graph Convolutional Network with High-order Chebyshev Approximation for Text Reasoning
https://aclanthology.org/2021.acl-long.513/
[ "Shuoran Jiang", "Qingcai Chen", "Xin Liu", "Baotian Hu", "Lisai Zhang" ]
Graph convolutional network (GCN) has become popular in various natural language processing (NLP) tasks with its superiority in long-term and non-consecutive word interactions. However, existing single-hop graph reasoning in GCN may miss some important non-consecutive dependencies. In this study, we define the spectral...
2021.acl-long.513
10.18653/v1/2021.acl-long.513
null
2106.05221
title_snapshot
2021.acl-long.514
Accelerating Text Communication via Abbreviated Sentence Input
https://aclanthology.org/2021.acl-long.514/
[ "Jiban Adhikary", "Jamie Berger", "Keith Vertanen" ]
Typing every character in a text message may require more time or effort than strictly necessary. Skipping spaces or other characters may be able to speed input and reduce a user’s physical input effort. This can be particularly important for people with motor impairments. In a large crowdsourced study, we found worker...
2021.acl-long.514
10.18653/v1/2021.acl-long.514
null
null
null
2021.acl-long.515
Regression Bugs Are In Your Model! Measuring, Reducing and Analyzing Regressions In NLP Model Updates
https://aclanthology.org/2021.acl-long.515/
[ "Yuqing Xie", "Yi-An Lai", "Yuanjun Xiong", "Yi Zhang", "Stefano Soatto" ]
Behavior of deep neural networks can be inconsistent between different versions. Regressions during model update are a common cause of concern that often over-weigh the benefits in accuracy or efficiency gain. This work focuses on quantifying, reducing and analyzing regression errors in the NLP model updates. Using neg...
2021.acl-long.515
10.18653/v1/2021.acl-long.515
null
2105.03048
title_snapshot
2021.acl-long.516
Detecting Propaganda Techniques in Memes
https://aclanthology.org/2021.acl-long.516/
[ "Dimitar Dimitrov", "Bishr Bin Ali", "Shaden Shaar", "Firoj Alam", "Fabrizio Silvestri", "Hamed Firooz", "Preslav Nakov", "Giovanni Da San Martino" ]
Propaganda can be defined as a form of communication that aims to influence the opinions or the actions of people towards a specific goal; this is achieved by means of well-defined rhetorical and psychological devices. Propaganda, in the form we know it today, can be dated back to the beginning of the 17th century. How...
2021.acl-long.516
10.18653/v1/2021.acl-long.516
null
2109.08013
title_snapshot
2021.acl-long.517
On the Efficacy of Adversarial Data Collection for Question Answering: Results from a Large-Scale Randomized Study
https://aclanthology.org/2021.acl-long.517/
[ "Divyansh Kaushik", "Douwe Kiela", "Zachary C. Lipton", "Wen-tau Yih" ]
In adversarial data collection (ADC), a human workforce interacts with a model in real time, attempting to produce examples that elicit incorrect predictions. Researchers hope that models trained on these more challenging datasets will rely less on superficial patterns, and thus be less brittle. However, despite ADC’s ...
2021.acl-long.517
10.18653/v1/2021.acl-long.517
null
2106.00872
title_snapshot
2021.acl-long.518
Learning Dense Representations of Phrases at Scale
https://aclanthology.org/2021.acl-long.518/
[ "Jinhyuk Lee", "Mujeen Sung", "Jaewoo Kang", "Danqi Chen" ]
Open-domain question answering can be reformulated as a phrase retrieval problem, without the need for processing documents on-demand during inference (Seo et al., 2019). However, current phrase retrieval models heavily depend on sparse representations and still underperform retriever-reader approaches. In this work, w...
2021.acl-long.518
10.18653/v1/2021.acl-long.518
null
2012.12624
title_snapshot
2021.acl-long.519
End-to-End Training of Neural Retrievers for Open-Domain Question Answering
https://aclanthology.org/2021.acl-long.519/
[ "Devendra Sachan", "Mostofa Patwary", "Mohammad Shoeybi", "Neel Kant", "Wei Ping", "William L. Hamilton", "Bryan Catanzaro" ]
Recent work on training neural retrievers for open-domain question answering (OpenQA) has employed both supervised and unsupervised approaches. However, it remains unclear how unsupervised and supervised methods can be used most effectively for neural retrievers. In this work, we systematically study retriever pre-trai...
2021.acl-long.519
10.18653/v1/2021.acl-long.519
null
2101.00408
title_snapshot
2021.acl-long.520
Question Answering Over Temporal Knowledge Graphs
https://aclanthology.org/2021.acl-long.520/
[ "Apoorv Saxena", "Soumen Chakrabarti", "Partha Talukdar" ]
Temporal Knowledge Graphs (Temporal KGs) extend regular Knowledge Graphs by providing temporal scopes (start and end times) on each edge in the KG. While Question Answering over KG (KGQA) has received some attention from the research community, QA over Temporal KGs (Temporal KGQA) is a relatively unexplored area. Lack ...
2021.acl-long.520
10.18653/v1/2021.acl-long.520
null
2106.01515
title_snapshot
2021.acl-long.521
Language Model Augmented Relevance Score
https://aclanthology.org/2021.acl-long.521/
[ "Ruibo Liu", "Jason Wei", "Soroush Vosoughi" ]
Although automated metrics are commonly used to evaluate NLG systems, they often correlate poorly with human judgements. Newer metrics such as BERTScore have addressed many weaknesses in prior metrics such as BLEU and ROUGE, which rely on n-gram matching. These newer methods, however, are still limited in that they do ...
2021.acl-long.521
10.18653/v1/2021.acl-long.521
null
2108.08485
title_snapshot
2021.acl-long.522
DExperts: Decoding-Time Controlled Text Generation with Experts and Anti-Experts
https://aclanthology.org/2021.acl-long.522/
[ "Alisa Liu", "Maarten Sap", "Ximing Lu", "Swabha Swayamdipta", "Chandra Bhagavatula", "Noah A. Smith", "Yejin Choi" ]
Despite recent advances in natural language generation, it remains challenging to control attributes of generated text. We propose DExperts: Decoding-time Experts, a decoding-time method for controlled text generation that combines a pretrained language model with “expert” LMs and/or “anti-expert” LMs in a product of e...
2021.acl-long.522
10.18653/v1/2021.acl-long.522
null
2105.03023
title_snapshot
2021.acl-long.523
Polyjuice: Generating Counterfactuals for Explaining, Evaluating, and Improving Models
https://aclanthology.org/2021.acl-long.523/
[ "Tongshuang Wu", "Marco Tulio Ribeiro", "Jeffrey Heer", "Daniel Weld" ]
While counterfactual examples are useful for analysis and training of NLP models, current generation methods either rely on manual labor to create very few counterfactuals, or only instantiate limited types of perturbations such as paraphrases or word substitutions. We present Polyjuice, a general-purpose counterfactua...
2021.acl-long.523
10.18653/v1/2021.acl-long.523
null
2101.00288
title_snapshot
2021.acl-long.524
Metaphor Generation with Conceptual Mappings
https://aclanthology.org/2021.acl-long.524/
[ "Kevin Stowe", "Tuhin Chakrabarty", "Nanyun Peng", "Smaranda Muresan", "Iryna Gurevych" ]
Generating metaphors is a difficult task as it requires understanding nuanced relationships between abstract concepts. In this paper, we aim to generate a metaphoric sentence given a literal expression by replacing relevant verbs. Guided by conceptual metaphor theory, we propose to control the generation process by enc...
2021.acl-long.524
10.18653/v1/2021.acl-long.524
null
2106.01228
title_snapshot
2021.acl-long.525
Learning Latent Structures for Cross Action Phrase Relations in Wet Lab Protocols
https://aclanthology.org/2021.acl-long.525/
[ "Chaitanya Kulkarni", "Jany Chan", "Eric Fosler-Lussier", "Raghu Machiraju" ]
Wet laboratory protocols (WLPs) are critical for conveying reproducible procedures in biological research. They are composed of instructions written in natural language describing the step-wise processing of materials by specific actions. This process flow description for reagents and materials synthesis in WLPs can be...
2021.acl-long.525
10.18653/v1/2021.acl-long.525
null
null
null
2021.acl-long.526
Multimodal Multi-Speaker Merger & Acquisition Financial Modeling: A New Task, Dataset, and Neural Baselines
https://aclanthology.org/2021.acl-long.526/
[ "Ramit Sawhney", "Mihir Goyal", "Prakhar Goel", "Puneet Mathur", "Rajiv Ratn Shah" ]
Risk prediction is an essential task in financial markets. Merger and Acquisition (M&A) calls provide key insights into the claims made by company executives about the restructuring of the financial firms. Extracting vocal and textual cues from M&A calls can help model the risk associated with such financial activities...
2021.acl-long.526
10.18653/v1/2021.acl-long.526
null
null
null
2021.acl-long.527
Mid-Air Hand Gestures for Post-Editing of Machine Translation
https://aclanthology.org/2021.acl-long.527/
[ "Rashad Albo Jamara", "Nico Herbig", "Antonio Krüger", "Josef van Genabith" ]
To translate large volumes of text in a globally connected world, more and more translators are integrating machine translation (MT) and post-editing (PE) into their translation workflows to generate publishable quality translations. While this process has been shown to save time and reduce errors, the task of translat...
2021.acl-long.527
10.18653/v1/2021.acl-long.527
null
null
null
2021.acl-long.528
Inter-GPS: Interpretable Geometry Problem Solving with Formal Language and Symbolic Reasoning
https://aclanthology.org/2021.acl-long.528/
[ "Pan Lu", "Ran Gong", "Shibiao Jiang", "Liang Qiu", "Siyuan Huang", "Xiaodan Liang", "Song-Chun Zhu" ]
Geometry problem solving has attracted much attention in the NLP community recently. The task is challenging as it requires abstract problem understanding and symbolic reasoning with axiomatic knowledge. However, current datasets are either small in scale or not publicly available. Thus, we construct a new large-scale ...
2021.acl-long.528
10.18653/v1/2021.acl-long.528
null
2105.04165
title_snapshot
2021.acl-long.529
Joint Verification and Reranking for Open Fact Checking Over Tables
https://aclanthology.org/2021.acl-long.529/
[ "Michael Sejr Schlichtkrull", "Vladimir Karpukhin", "Barlas Oguz", "Mike Lewis", "Wen-tau Yih", "Sebastian Riedel" ]
Structured information is an important knowledge source for automatic verification of factual claims. Nevertheless, the majority of existing research into this task has focused on textual data, and the few recent inquiries into structured data have been for the closed-domain setting where appropriate evidence for each ...
2021.acl-long.529
10.18653/v1/2021.acl-long.529
null
2012.15115
title_snapshot
2021.acl-long.530
Evaluation of Thematic Coherence in Microblogs
https://aclanthology.org/2021.acl-long.530/
[ "Iman Munire Bilal", "Bo Wang", "Maria Liakata", "Rob Procter", "Adam Tsakalidis" ]
Collecting together microblogs representing opinions about the same topics within the same timeframe is useful to a number of different tasks and practitioners. A major question is how to evaluate the quality of such thematic clusters. Here we create a corpus of microblog clusters from three different domains and time ...
2021.acl-long.530
10.18653/v1/2021.acl-long.530
null
2106.15971
title_snapshot
2021.acl-long.531
Neural semi-Markov CRF for Monolingual Word Alignment
https://aclanthology.org/2021.acl-long.531/
[ "Wuwei Lan", "Chao Jiang", "Wei Xu" ]
Monolingual word alignment is important for studying fine-grained editing operations (i.e., deletion, addition, and substitution) in text-to-text generation tasks, such as paraphrase generation, text simplification, neutralizing biased language, etc. In this paper, we present a novel neural semi-Markov CRF alignment mo...
2021.acl-long.531
10.18653/v1/2021.acl-long.531
null
2106.02569
title_snapshot
2021.acl-long.532
Privacy at Scale: Introducing the PrivaSeer Corpus of Web Privacy Policies
https://aclanthology.org/2021.acl-long.532/
[ "Mukund Srinath", "Shomir Wilson", "C Lee Giles" ]
Organisations disclose their privacy practices by posting privacy policies on their websites. Even though internet users often care about their digital privacy, they usually do not read privacy policies, since understanding them requires a significant investment of time and effort. Natural language processing has been ...
2021.acl-long.532
10.18653/v1/2021.acl-long.532
null
2004.11131
title_snapshot
2021.acl-long.533
The statistical advantage of automatic NLG metrics at the system level
https://aclanthology.org/2021.acl-long.533/
[ "Johnny Wei", "Robin Jia" ]
Estimating the expected output quality of generation systems is central to NLG. This paper qualifies the notion that automatic metrics are not as good as humans in estimating system-level quality. Statistically, humans are unbiased, high variance estimators, while metrics are biased, low variance estimators. We compare...
2021.acl-long.533
10.18653/v1/2021.acl-long.533
null
2105.12437
title_snapshot
2021.acl-long.534
Are Missing Links Predictable? An Inferential Benchmark for Knowledge Graph Completion
https://aclanthology.org/2021.acl-long.534/
[ "Yixin Cao", "Xiang Ji", "Xin Lv", "Juanzi Li", "Yonggang Wen", "Hanwang Zhang" ]
We present InferWiki, a Knowledge Graph Completion (KGC) dataset that improves upon existing benchmarks in inferential ability, assumptions, and patterns. First, each testing sample is predictable with supportive data in the training set. To ensure it, we propose to utilize rule-guided train/test generation, instead of...
2021.acl-long.534
10.18653/v1/2021.acl-long.534
null
2108.01387
title_snapshot
2021.acl-long.535
ConvoSumm: Conversation Summarization Benchmark and Improved Abstractive Summarization with Argument Mining
https://aclanthology.org/2021.acl-long.535/
[ "Alexander Fabbri", "Faiaz Rahman", "Imad Rizvi", "Borui Wang", "Haoran Li", "Yashar Mehdad", "Dragomir Radev" ]
While online conversations can cover a vast amount of information in many different formats, abstractive text summarization has primarily focused on modeling solely news articles. This research gap is due, in part, to the lack of standardized datasets for summarizing online discussions. To address this gap, we design a...
2021.acl-long.535
10.18653/v1/2021.acl-long.535
null
2106.00829
title_snapshot
2021.acl-long.536
Improving Factual Consistency of Abstractive Summarization via Question Answering
https://aclanthology.org/2021.acl-long.536/
[ "Feng Nan", "Cicero Nogueira dos Santos", "Henghui Zhu", "Patrick Ng", "Kathleen McKeown", "Ramesh Nallapati", "Dejiao Zhang", "Zhiguo Wang", "Andrew O. Arnold", "Bing Xiang" ]
A commonly observed problem with the state-of-the art abstractive summarization models is that the generated summaries can be factually inconsistent with the input documents. The fact that automatic summarization may produce plausible-sounding yet inaccurate summaries is a major concern that limits its wide application...
2021.acl-long.536
10.18653/v1/2021.acl-long.536
null
2105.04623
title_snapshot
2021.acl-long.537
EmailSum: Abstractive Email Thread Summarization
https://aclanthology.org/2021.acl-long.537/
[ "Shiyue Zhang", "Asli Celikyilmaz", "Jianfeng Gao", "Mohit Bansal" ]
Recent years have brought about an interest in the challenging task of summarizing conversation threads (meetings, online discussions, etc.). Such summaries help analysis of the long text to quickly catch up with the decisions made and thus improve our work or communication efficiency. To spur research in thread summar...
2021.acl-long.537
10.18653/v1/2021.acl-long.537
null
2107.14691
title_snapshot
2021.acl-long.538
Cross-Lingual Abstractive Summarization with Limited Parallel Resources
https://aclanthology.org/2021.acl-long.538/
[ "Yu Bai", "Yang Gao", "Heyan Huang" ]
Parallel cross-lingual summarization data is scarce, requiring models to better use the limited available cross-lingual resources. Existing methods to do so often adopt sequence-to-sequence networks with multi-task frameworks. Such approaches apply multiple decoders, each of which is utilized for a specific task. Howev...
2021.acl-long.538
10.18653/v1/2021.acl-long.538
null
2105.13648
title_snapshot
2021.acl-long.539
Dissecting Generation Modes for Abstractive Summarization Models via Ablation and Attribution
https://aclanthology.org/2021.acl-long.539/
[ "Jiacheng Xu", "Greg Durrett" ]
Despite the prominence of neural abstractive summarization models, we know little about how they actually form summaries and how to understand where their decisions come from. We propose a two-step method to interpret summarization model decisions. We first analyze the model’s behavior by ablating the full model to cat...
2021.acl-long.539
10.18653/v1/2021.acl-long.539
null
2106.01518
title_snapshot
2021.acl-long.540
Learning Prototypical Functions for Physical Artifacts
https://aclanthology.org/2021.acl-long.540/
[ "Tianyu Jiang", "Ellen Riloff" ]
Humans create things for a reason. Ancient people created spears for hunting, knives for cutting meat, pots for preparing food, etc. The prototypical function of a physical artifact is a kind of commonsense knowledge that we rely on to understand natural language. For example, if someone says “She borrowed the book” th...
2021.acl-long.540
10.18653/v1/2021.acl-long.540
null
null
null
2021.acl-long.541
Verb Knowledge Injection for Multilingual Event Processing
https://aclanthology.org/2021.acl-long.541/
[ "Olga Majewska", "Ivan Vulić", "Goran Glavaš", "Edoardo Maria Ponti", "Anna Korhonen" ]
Linguistic probing of pretrained Transformer-based language models (LMs) revealed that they encode a range of syntactic and semantic properties of a language. However, they are still prone to fall back on superficial cues and simple heuristics to solve downstream tasks, rather than leverage deeper linguistic informatio...
2021.acl-long.541
10.18653/v1/2021.acl-long.541
null
2012.15421
title_snapshot
2021.acl-long.542
Dynamic Contextualized Word Embeddings
https://aclanthology.org/2021.acl-long.542/
[ "Valentin Hofmann", "Janet Pierrehumbert", "Hinrich Schütze" ]
Static word embeddings that represent words by a single vector cannot capture the variability of word meaning in different linguistic and extralinguistic contexts. Building on prior work on contextualized and dynamic word embeddings, we introduce dynamic contextualized word embeddings that represent words as a function...
2021.acl-long.542
10.18653/v1/2021.acl-long.542
null
2010.12684
title_snapshot
2021.acl-long.543
Lexical Semantic Change Discovery
https://aclanthology.org/2021.acl-long.543/
[ "Sinan Kurtyigit", "Maike Park", "Dominik Schlechtweg", "Jonas Kuhn", "Sabine Schulte im Walde" ]
While there is a large amount of research in the field of Lexical Semantic Change Detection, only few approaches go beyond a standard benchmark evaluation of existing models. In this paper, we propose a shift of focus from change detection to change discovery, i.e., discovering novel word senses over time from the full...
2021.acl-long.543
10.18653/v1/2021.acl-long.543
null
2106.03111
title_snapshot
2021.acl-long.544
The R-U-A-Robot Dataset: Helping Avoid Chatbot Deception by Detecting User Questions About Human or Non-Human Identity
https://aclanthology.org/2021.acl-long.544/
[ "David Gros", "Yu Li", "Zhou Yu" ]
Humans are increasingly interacting with machines through language, sometimes in contexts where the user may not know they are talking to a machine (like over the phone or a text chatbot). We aim to understand how system designers and researchers might allow their systems to confirm its non-human identity. We collect o...
2021.acl-long.544
10.18653/v1/2021.acl-long.544
null
2106.02692
title_snapshot
2021.acl-long.545
Using Meta-Knowledge Mined from Identifiers to Improve Intent Recognition in Conversational Systems
https://aclanthology.org/2021.acl-long.545/
[ "Claudio Pinhanez", "Paulo Cavalin", "Victor Henrique Alves Ribeiro", "Ana Appel", "Heloisa Candello", "Julio Nogima", "Mauro Pichiliani", "Melina Guerra", "Maira de Bayser", "Gabriel Malfatti", "Henrique Ferreira" ]
In this paper we explore the improvement of intent recognition in conversational systems by the use of meta-knowledge embedded in intent identifiers. Developers often include such knowledge, structure as taxonomies, in the documentation of chatbots. By using neuro-symbolic algorithms to incorporate those taxonomies int...
2021.acl-long.545
10.18653/v1/2021.acl-long.545
null
2012.09005
title_judge
2021.acl-long.546
Space Efficient Context Encoding for Non-Task-Oriented Dialogue Generation with Graph Attention Transformer
https://aclanthology.org/2021.acl-long.546/
[ "Fabian Galetzka", "Jewgeni Rose", "David Schlangen", "Jens Lehmann" ]
To improve the coherence and knowledge retrieval capabilities of non-task-oriented dialogue systems, recent Transformer-based models aim to integrate fixed background context. This often comes in the form of knowledge graphs, and the integration is done by creating pseudo utterances through paraphrasing knowledge tripl...
2021.acl-long.546
10.18653/v1/2021.acl-long.546
null
null
null
2021.acl-long.547
DialogueCRN: Contextual Reasoning Networks for Emotion Recognition in Conversations
https://aclanthology.org/2021.acl-long.547/
[ "Dou Hu", "Lingwei Wei", "Xiaoyong Huai" ]
Emotion Recognition in Conversations (ERC) has gained increasing attention for developing empathetic machines. Recently, many approaches have been devoted to perceiving conversational context by deep learning models. However, these approaches are insufficient in understanding the context due to lacking the ability to e...
2021.acl-long.547
10.18653/v1/2021.acl-long.547
null
2106.01978
title_snapshot
2021.acl-long.548
Cross-replication Reliability - An Empirical Approach to Interpreting Inter-rater Reliability
https://aclanthology.org/2021.acl-long.548/
[ "Ka Wong", "Praveen Paritosh", "Lora Aroyo" ]
When collecting annotations and labeled data from humans, a standard practice is to use inter-rater reliability (IRR) as a measure of data goodness (Hallgren, 2012). Metrics such as Krippendorff’s alpha or Cohen’s kappa are typically required to be above a threshold of 0.6 (Landis and Koch, 1977). These absolute thresh...
2021.acl-long.548
10.18653/v1/2021.acl-long.548
null
2106.07393
title_snapshot
2021.acl-long.549
TIMEDIAL: Temporal Commonsense Reasoning in Dialog
https://aclanthology.org/2021.acl-long.549/
[ "Lianhui Qin", "Aditya Gupta", "Shyam Upadhyay", "Luheng He", "Yejin Choi", "Manaal Faruqui" ]
Everyday conversations require understanding everyday events, which in turn, requires understanding temporal commonsense concepts interwoven with those events. Despite recent progress with massive pre-trained language models (LMs) such as T5 and GPT-3, their capability of temporal reasoning in dialogs remains largely u...
2021.acl-long.549
10.18653/v1/2021.acl-long.549
null
2106.04571
title_snapshot
2021.acl-long.550
RAW-C: Relatedness of Ambiguous Words in Context (A New Lexical Resource for English)
https://aclanthology.org/2021.acl-long.550/
[ "Sean Trott", "Benjamin Bergen" ]
Most words are ambiguous—-i.e., they convey distinct meanings in different contexts—-and even the meanings of unambiguous words are context-dependent. Both phenomena present a challenge for NLP. Recently, the advent of contextualized word embeddings has led to success on tasks involving lexical ambiguity, such as Word ...
2021.acl-long.550
10.18653/v1/2021.acl-long.550
null
2105.13266
title_snapshot
2021.acl-long.551
ARBERT & MARBERT: Deep Bidirectional Transformers for Arabic
https://aclanthology.org/2021.acl-long.551/
[ "Muhammad Abdul-Mageed", "AbdelRahim Elmadany", "El Moatez Billah Nagoudi" ]
Pre-trained language models (LMs) are currently integral to many natural language processing systems. Although multilingual LMs were also introduced to serve many languages, these have limitations such as being costly at inference time and the size and diversity of non-English data involved in their pre-training. We re...
2021.acl-long.551
10.18653/v1/2021.acl-long.551
null
2101.01785
title_snapshot
2021.acl-long.552
Improving Paraphrase Detection with the Adversarial Paraphrasing Task
https://aclanthology.org/2021.acl-long.552/
[ "Animesh Nighojkar", "John Licato" ]
If two sentences have the same meaning, it should follow that they are equivalent in their inferential properties, i.e., each sentence should textually entail the other. However, many paraphrase datasets currently in widespread use rely on a sense of paraphrase based on word overlap and syntax. Can we teach them instea...
2021.acl-long.552
10.18653/v1/2021.acl-long.552
null
2106.07691
title_snapshot
2021.acl-long.553
ADEPT: An Adjective-Dependent Plausibility Task
https://aclanthology.org/2021.acl-long.553/
[ "Ali Emami", "Ian Porada", "Alexandra Olteanu", "Kaheer Suleman", "Adam Trischler", "Jackie Chi Kit Cheung" ]
A false contract is more likely to be rejected than a contract is, yet a false key is less likely than a key to open doors. While correctly interpreting and assessing the effects of such adjective-noun pairs (e.g., false key) on the plausibility of given events (e.g., opening doors) underpins many natural language unde...
2021.acl-long.553
10.18653/v1/2021.acl-long.553
null
null
null
2021.acl-long.554
ReadOnce Transformers: Reusable Representations of Text for Transformers
https://aclanthology.org/2021.acl-long.554/
[ "Shih-Ting Lin", "Ashish Sabharwal", "Tushar Khot" ]
We present ReadOnce Transformers, an approach to convert a transformer-based model into one that can build an information-capturing, task-independent, and compressed representation of text. The resulting representation is reusable across different examples and tasks, thereby requiring a document shared across many exam...
2021.acl-long.554
10.18653/v1/2021.acl-long.554
null
2010.12854
title_snapshot
2021.acl-long.555
Conditional Generation of Temporally-ordered Event Sequences
https://aclanthology.org/2021.acl-long.555/
[ "Shih-Ting Lin", "Nathanael Chambers", "Greg Durrett" ]
Models of narrative schema knowledge have proven useful for a range of event-related tasks, but they typically do not capture the temporal relationships between events. We propose a single model that addresses both temporal ordering, sorting given events into the order they occurred, and event infilling, predicting new...
2021.acl-long.555
10.18653/v1/2021.acl-long.555
null
2012.15786
title_snapshot
2021.acl-long.556
Hate Speech Detection Based on Sentiment Knowledge Sharing
https://aclanthology.org/2021.acl-long.556/
[ "Xianbing Zhou", "Yang Yong", "Xiaochao Fan", "Ge Ren", "Yunfeng Song", "Yufeng Diao", "Liang Yang", "Hongfei Lin" ]
The wanton spread of hate speech on the internet brings great harm to society and families. It is urgent to establish and improve automatic detection and active avoidance mechanisms for hate speech. While there exist methods for hate speech detection, they stereotype words and hence suffer from inherently biased traini...
2021.acl-long.556
10.18653/v1/2021.acl-long.556
null
null
null
2021.acl-long.557
Transition-based Bubble Parsing: Improvements on Coordination Structure Prediction
https://aclanthology.org/2021.acl-long.557/
[ "Tianze Shi", "Lillian Lee" ]
We propose a transition-based bubble parser to perform coordination structure identification and dependency-based syntactic analysis simultaneously. Bubble representations were proposed in the formal linguistics literature decades ago; they enhance dependency trees by encoding coordination boundaries and internal relat...
2021.acl-long.557
10.18653/v1/2021.acl-long.557
null
2107.06905
title_snapshot
2021.acl-long.558
SpanNER: Named Entity Re-/Recognition as Span Prediction
https://aclanthology.org/2021.acl-long.558/
[ "Jinlan Fu", "Xuanjing Huang", "Pengfei Liu" ]
Recent years have seen the paradigm shift of Named Entity Recognition (NER) systems from sequence labeling to span prediction. Despite its preliminary effectiveness, the span prediction model’s architectural bias has not been fully understood. In this paper, we first investigate the strengths and weaknesses when the sp...
2021.acl-long.558
10.18653/v1/2021.acl-long.558
null
2106.00641
title_snapshot
2021.acl-long.559
StructFormer: Joint Unsupervised Induction of Dependency and Constituency Structure from Masked Language Modeling
https://aclanthology.org/2021.acl-long.559/
[ "Yikang Shen", "Yi Tay", "Che Zheng", "Dara Bahri", "Donald Metzler", "Aaron Courville" ]
There are two major classes of natural language grammars — the dependency grammar that models one-to-one correspondences between words and the constituency grammar that models the assembly of one or several corresponded words. While previous unsupervised parsing methods mostly focus on only inducing one class of gramma...
2021.acl-long.559
10.18653/v1/2021.acl-long.559
null
2012.00857
title_snapshot
2021.acl-long.560
Language Embeddings for Typology and Cross-lingual Transfer Learning
https://aclanthology.org/2021.acl-long.560/
[ "Dian Yu", "Taiqi He", "Kenji Sagae" ]
Cross-lingual language tasks typically require a substantial amount of annotated data or parallel translation data. We explore whether language representations that capture relationships among languages can be learned and subsequently leveraged in cross-lingual tasks without the use of parallel data. We generate dense ...
2021.acl-long.560
10.18653/v1/2021.acl-long.560
null
2106.02082
title_snapshot
2021.acl-long.561
Can Sequence-to-Sequence Models Crack Substitution Ciphers?
https://aclanthology.org/2021.acl-long.561/
[ "Nada Aldarrab", "Jonathan May" ]
Decipherment of historical ciphers is a challenging problem. The language of the target plaintext might be unknown, and ciphertext can have a lot of noise. State-of-the-art decipherment methods use beam search and a neural language model to score candidate plaintext hypotheses for a given cipher, assuming the plaintext...
2021.acl-long.561
10.18653/v1/2021.acl-long.561
null
2012.15229
title_snapshot
2021.acl-long.562
Beyond Noise: Mitigating the Impact of Fine-grained Semantic Divergences on Neural Machine Translation
https://aclanthology.org/2021.acl-long.562/
[ "Eleftheria Briakou", "Marine Carpuat" ]
While it has been shown that Neural Machine Translation (NMT) is highly sensitive to noisy parallel training samples, prior work treats all types of mismatches between source and target as noise. As a result, it remains unclear how samples that are mostly equivalent but contain a small number of semantically divergent ...
2021.acl-long.562
10.18653/v1/2021.acl-long.562
null
2105.15087
title_snapshot
2021.acl-long.563
Discriminative Reranking for Neural Machine Translation
https://aclanthology.org/2021.acl-long.563/
[ "Ann Lee", "Michael Auli", "Marc’Aurelio Ranzato" ]
Reranking models enable the integration of rich features to select a better output hypothesis within an n-best list or lattice. These models have a long history in NLP, and we revisit discriminative reranking for modern neural machine translation models by training a large transformer architecture. This takes as input ...
2021.acl-long.563
10.18653/v1/2021.acl-long.563
null
null
null
2021.acl-long.564
Mind Your Outliers! Investigating the Negative Impact of Outliers on Active Learning for Visual Question Answering
https://aclanthology.org/2021.acl-long.564/
[ "Siddharth Karamcheti", "Ranjay Krishna", "Li Fei-Fei", "Christopher Manning" ]
Active learning promises to alleviate the massive data needs of supervised machine learning: it has successfully improved sample efficiency by an order of magnitude on traditional tasks like topic classification and object recognition. However, we uncover a striking contrast to this promise: across 5 models and 4 datas...
2021.acl-long.564
10.18653/v1/2021.acl-long.564
Outstanding Paper
2107.02331
title_snapshot
2021.acl-long.565
All That’s ‘Human’ Is Not Gold: Evaluating Human Evaluation of Generated Text
https://aclanthology.org/2021.acl-long.565/
[ "Elizabeth Clark", "Tal August", "Sofia Serrano", "Nikita Haduong", "Suchin Gururangan", "Noah A. Smith" ]
Human evaluations are typically considered the gold standard in natural language generation, but as models’ fluency improves, how well can evaluators detect and judge machine-generated text? We run a study assessing non-experts’ ability to distinguish between human- and machine-authored text (GPT2 and GPT3) in three do...
2021.acl-long.565
10.18653/v1/2021.acl-long.565
Outstanding Paper
2107.00061
title_snapshot
2021.acl-long.566
Scientific Credibility of Machine Translation Research: A Meta-Evaluation of 769 Papers
https://aclanthology.org/2021.acl-long.566/
[ "Benjamin Marie", "Atsushi Fujita", "Raphael Rubino" ]
This paper presents the first large-scale meta-evaluation of machine translation (MT). We annotated MT evaluations conducted in 769 research papers published from 2010 to 2020. Our study shows that practices for automatic MT evaluation have dramatically changed during the past decade and follow concerning trends. An in...
2021.acl-long.566
10.18653/v1/2021.acl-long.566
Outstanding Paper
2106.15195
title_snapshot
2021.acl-long.567
Neural Machine Translation with Monolingual Translation Memory
https://aclanthology.org/2021.acl-long.567/
[ "Deng Cai", "Yan Wang", "Huayang Li", "Wai Lam", "Lemao Liu" ]
Prior work has proved that Translation Memory (TM) can boost the performance of Neural Machine Translation (NMT). In contrast to existing work that uses bilingual corpus as TM and employs source-side similarity search for memory retrieval, we propose a new framework that uses monolingual memory and performs learnable m...
2021.acl-long.567
10.18653/v1/2021.acl-long.567
Outstanding Paper
2105.11269
title_snapshot
2021.acl-long.568
Intrinsic Dimensionality Explains the Effectiveness of Language Model Fine-Tuning
https://aclanthology.org/2021.acl-long.568/
[ "Armen Aghajanyan", "Sonal Gupta", "Luke Zettlemoyer" ]
Although pretrained language models can be fine-tuned to produce state-of-the-art results for a very wide range of language understanding tasks, the dynamics of this process are not well understood, especially in the low data regime. Why can we use relatively vanilla gradient descent algorithms (e.g., without strong re...
2021.acl-long.568
10.18653/v1/2021.acl-long.568
Outstanding Paper
2012.13255
title_snapshot
2021.acl-long.569
UnNatural Language Inference
https://aclanthology.org/2021.acl-long.569/
[ "Koustuv Sinha", "Prasanna Parthasarathi", "Joelle Pineau", "Adina Williams" ]
Recent investigations into the inner-workings of state-of-the-art large-scale pre-trained Transformer-based Natural Language Understanding (NLU) models indicate that they appear to understand human-like syntax, at least to some extent. We provide novel evidence that complicates this claim: we find that state-of-the-art...
2021.acl-long.569
10.18653/v1/2021.acl-long.569
Outstanding Paper
2101.00010
title_snapshot
2021.acl-long.570
Including Signed Languages in Natural Language Processing
https://aclanthology.org/2021.acl-long.570/
[ "Kayo Yin", "Amit Moryossef", "Julie Hochgesang", "Yoav Goldberg", "Malihe Alikhani" ]
Signed languages are the primary means of communication for many deaf and hard of hearing individuals. Since signed languages exhibit all the fundamental linguistic properties of natural language, we believe that tools and theories of Natural Language Processing (NLP) are crucial towards its modeling. However, existing...
2021.acl-long.570
10.18653/v1/2021.acl-long.570
Best Theme Paper
2105.05222
title_snapshot
2021.acl-long.571
Vocabulary Learning via Optimal Transport for Neural Machine Translation
https://aclanthology.org/2021.acl-long.571/
[ "Jingjing Xu", "Hao Zhou", "Chun Gan", "Zaixiang Zheng", "Lei Li" ]
The choice of token vocabulary affects the performance of machine translation. This paper aims to figure out what is a good vocabulary and whether we can find the optimal vocabulary without trial training. To answer these questions, we first provide an alternative understanding of vocabulary from the perspective of inf...
2021.acl-long.571
10.18653/v1/2021.acl-long.571
Best Paper
2012.15671
title_snapshot
2021.acl-short.1
Catchphrase: Automatic Detection of Cultural References
https://aclanthology.org/2021.acl-short.1/
[ "Nir Sweed", "Dafna Shahaf" ]
A snowclone is a customizable phrasal template that can be realized in multiple, instantly recognized variants. For example, “* is the new *" (Orange is the new black, 40 is the new 30). Snowclones are extensively used in social media. In this paper, we study snowclones originating from pop-culture quotes; our goal is ...
2021.acl-short.1
10.18653/v1/2021.acl-short.1
null
2106.04830
title_snapshot
2021.acl-short.2
On Training Instance Selection for Few-Shot Neural Text Generation
https://aclanthology.org/2021.acl-short.2/
[ "Ernie Chang", "Xiaoyu Shen", "Hui-Syuan Yeh", "Vera Demberg" ]
Large-scale pretrained language models have led to dramatic improvements in text generation. Impressive performance can be achieved by finetuning only on a small number of instances (few-shot setting). Nonetheless, almost all previous work simply applies random sampling to select the few-shot training instances. Little...
2021.acl-short.2
10.18653/v1/2021.acl-short.2
null
2107.03176
title_snapshot
2021.acl-short.3
Coreference Resolution without Span Representations
https://aclanthology.org/2021.acl-short.3/
[ "Yuval Kirstain", "Ori Ram", "Omer Levy" ]
The introduction of pretrained language models has reduced many complex task-specific NLP models to simple lightweight layers. An exception to this trend is coreference resolution, where a sophisticated task-specific model is appended to a pretrained transformer encoder. While highly effective, the model has a very lar...
2021.acl-short.3
10.18653/v1/2021.acl-short.3
null
2101.00434
title_snapshot
2021.acl-short.4
Enhancing Entity Boundary Detection for Better Chinese Named Entity Recognition
https://aclanthology.org/2021.acl-short.4/
[ "Chun Chen", "Fang Kong" ]
In comparison with English, due to the lack of explicit word boundary and tenses information, Chinese Named Entity Recognition (NER) is much more challenging. In this paper, we propose a boundary enhanced approach for better Chinese NER. In particular, our approach enhances the boundary information from two perspective...
2021.acl-short.4
10.18653/v1/2021.acl-short.4
null
null
null
2021.acl-short.5
Difficulty-Aware Machine Translation Evaluation
https://aclanthology.org/2021.acl-short.5/
[ "Runzhe Zhan", "Xuebo Liu", "Derek F. Wong", "Lidia S. Chao" ]
The high-quality translation results produced by machine translation (MT) systems still pose a huge challenge for automatic evaluation. Current MT evaluation pays the same attention to each sentence component, while the questions of real-world examinations (e.g., university examinations) have different difficulties and...
2021.acl-short.5
10.18653/v1/2021.acl-short.5
null
2107.14402
title_snapshot
2021.acl-short.6
Uncertainty and Surprisal Jointly Deliver the Punchline: Exploiting Incongruity-Based Features for Humor Recognition
https://aclanthology.org/2021.acl-short.6/
[ "Yubo Xie", "Junze Li", "Pearl Pu" ]
Humor recognition has been widely studied as a text classification problem using data-driven approaches. However, most existing work does not examine the actual joke mechanism to understand humor. We break down any joke into two distinct components: the set-up and the punchline, and further explore the special relation...
2021.acl-short.6
10.18653/v1/2021.acl-short.6
null
2012.12007
title_snapshot
2021.acl-short.7
Counterfactuals to Control Latent Disentangled Text Representations for Style Transfer
https://aclanthology.org/2021.acl-short.7/
[ "Sharmila Reddy Nangi", "Niyati Chhaya", "Sopan Khosla", "Nikhil Kaushik", "Harshit Nyati" ]
Disentanglement of latent representations into content and style spaces has been a commonly employed method for unsupervised text style transfer. These techniques aim to learn the disentangled representations and tweak them to modify the style of a sentence. In this paper, we propose a counterfactual-based method to mo...
2021.acl-short.7
10.18653/v1/2021.acl-short.7
null
null
null
2021.acl-short.8
Attention Flows are Shapley Value Explanations
https://aclanthology.org/2021.acl-short.8/
[ "Kawin Ethayarajh", "Dan Jurafsky" ]
Shapley Values, a solution to the credit assignment problem in cooperative game theory, are a popular type of explanation in machine learning, having been used to explain the importance of features, embeddings, and even neurons. In NLP, however, leave-one-out and attention-based explanations still predominate. Can we d...
2021.acl-short.8
10.18653/v1/2021.acl-short.8
null
2105.14652
title_snapshot
2021.acl-short.9
Video Paragraph Captioning as a Text Summarization Task
https://aclanthology.org/2021.acl-short.9/
[ "Hui Liu", "Xiaojun Wan" ]
Video paragraph captioning aims to generate a set of coherent sentences to describe a video that contains several events. Most previous methods simplify this task by using ground-truth event segments. In this work, we propose a novel framework by taking this task as a text summarization task. We first generate lots of ...
2021.acl-short.9
10.18653/v1/2021.acl-short.9
null
null
null
2021.acl-short.10
Are VQA Systems RAD? Measuring Robustness to Augmented Data with Focused Interventions
https://aclanthology.org/2021.acl-short.10/
[ "Daniel Rosenberg", "Itai Gat", "Amir Feder", "Roi Reichart" ]
Deep learning algorithms have shown promising results in visual question answering (VQA) tasks, but a more careful look reveals that they often do not understand the rich signal they are being fed with. To understand and better measure the generalization capabilities of VQA systems, we look at their robustness to count...
2021.acl-short.10
10.18653/v1/2021.acl-short.10
null
2106.04484
title_snapshot
2021.acl-short.11
How Helpful is Inverse Reinforcement Learning for Table-to-Text Generation?
https://aclanthology.org/2021.acl-short.11/
[ "Sayan Ghosh", "Zheng Qi", "Snigdha Chaturvedi", "Shashank Srivastava" ]
Existing approaches for the Table-to-Text task suffer from issues such as missing information, hallucination and repetition. Many approaches to this problem use Reinforcement Learning (RL), which maximizes a single manually defined reward, such as BLEU. In this work, we instead pose the Table-to-Text task as Inverse Re...
2021.acl-short.11
10.18653/v1/2021.acl-short.11
null
null
null
2021.acl-short.12
Automatic Fake News Detection: Are Models Learning to Reason?
https://aclanthology.org/2021.acl-short.12/
[ "Casper Hansen", "Christian Hansen", "Lucas Chaves Lima" ]
Most fact checking models for automatic fake news detection are based on reasoning: given a claim with associated evidence, the models aim to estimate the claim veracity based on the supporting or refuting content within the evidence. When these models perform well, it is generally assumed to be due to the models havin...
2021.acl-short.12
10.18653/v1/2021.acl-short.12
null
2105.07698
title_snapshot
2021.acl-short.13
Saying No is An Art: Contextualized Fallback Responses for Unanswerable Dialogue Queries
https://aclanthology.org/2021.acl-short.13/
[ "Ashish Shrivastava", "Kaustubh Dhole", "Abhinav Bhatt", "Sharvani Raghunath" ]
Despite end-to-end neural systems making significant progress in the last decade for task-oriented as well as chit-chat based dialogue systems, most dialogue systems rely on hybrid approaches which use a combination of rule-based, retrieval and generative approaches for generating a set of ranked responses. Such dialog...
2021.acl-short.13
10.18653/v1/2021.acl-short.13
null
2012.01873
title_snapshot
2021.acl-short.14
N-Best ASR Transformer: Enhancing SLU Performance using Multiple ASR Hypotheses
https://aclanthology.org/2021.acl-short.14/
[ "Karthik Ganesan", "Pakhi Bamdev", "Jaivarsan B", "Amresh Venugopal", "Abhinav Tushar" ]
Spoken Language Understanding (SLU) systems parse speech into semantic structures like dialog acts and slots. This involves the use of an Automatic Speech Recognizer (ASR) to transcribe speech into multiple text alternatives (hypotheses). Transcription errors, ordinary in ASRs, impact downstream SLU performance negativ...
2021.acl-short.14
10.18653/v1/2021.acl-short.14
null
2106.06519
title_snapshot
2021.acl-short.15
Gender bias amplification during Speed-Quality optimization in Neural Machine Translation
https://aclanthology.org/2021.acl-short.15/
[ "Adithya Renduchintala", "Denise Diaz", "Kenneth Heafield", "Xian Li", "Mona Diab" ]
Is bias amplified when neural machine translation (NMT) models are optimized for speed and evaluated on generic test sets using BLEU? We investigate architectures and techniques commonly used to speed up decoding in Transformer-based models, such as greedy search, quantization, average attention networks (AANs) and sha...
2021.acl-short.15
10.18653/v1/2021.acl-short.15
null
2106.00169
title_snapshot
2021.acl-short.16
Machine Translation into Low-resource Language Varieties
https://aclanthology.org/2021.acl-short.16/
[ "Sachin Kumar", "Antonios Anastasopoulos", "Shuly Wintner", "Yulia Tsvetkov" ]
State-of-the-art machine translation (MT) systems are typically trained to generate “standard” target language; however, many languages have multiple varieties (regional varieties, dialects, sociolects, non-native varieties) that are different from the standard language. Such varieties are often low-resource, and hence...
2021.acl-short.16
10.18653/v1/2021.acl-short.16
null
2106.06797
title_snapshot
2021.acl-short.17
Is Sparse Attention more Interpretable?
https://aclanthology.org/2021.acl-short.17/
[ "Clara Meister", "Stefan Lazov", "Isabelle Augenstein", "Ryan Cotterell" ]
Sparse attention has been claimed to increase model interpretability under the assumption that it highlights influential inputs. Yet the attention distribution is typically over representations internal to the model rather than the inputs themselves, suggesting this assumption may not have merit. We build on the recent...
2021.acl-short.17
10.18653/v1/2021.acl-short.17
null
2106.01087
title_snapshot
2021.acl-short.18
The Case for Translation-Invariant Self-Attention in Transformer-Based Language Models
https://aclanthology.org/2021.acl-short.18/
[ "Ulme Wennberg", "Gustav Eje Henter" ]
Mechanisms for encoding positional information are central for transformer-based language models. In this paper, we analyze the position embeddings of existing language models, finding strong evidence of translation invariance, both for the embeddings themselves and for their effect on self-attention. The degree of tra...
2021.acl-short.18
10.18653/v1/2021.acl-short.18
null
2106.01950
title_snapshot
2021.acl-short.19
Relative Importance in Sentence Processing
https://aclanthology.org/2021.acl-short.19/
[ "Nora Hollenstein", "Lisa Beinborn" ]
Determining the relative importance of the elements in a sentence is a key factor for effortless natural language understanding. For human language processing, we can approximate patterns of relative importance by measuring reading fixations using eye-tracking technology. In neural language models, gradient-based salie...
2021.acl-short.19
10.18653/v1/2021.acl-short.19
null
2106.03471
title_snapshot
2021.acl-short.20
Doing Good or Doing Right? Exploring the Weakness of Commonsense Causal Reasoning Models
https://aclanthology.org/2021.acl-short.20/
[ "Mingyue Han", "Yinglin Wang" ]
Pretrained language models (PLM) achieve surprising performance on the Choice of Plausible Alternatives (COPA) task. However, whether PLMs have truly acquired the ability of causal reasoning remains a question. In this paper, we investigate the problem of semantic similarity bias and reveal the vulnerability of current...
2021.acl-short.20
10.18653/v1/2021.acl-short.20
null
2107.01791
title_snapshot
2021.acl-short.21
AND does not mean OR: Using Formal Languages to Study Language Models’ Representations
https://aclanthology.org/2021.acl-short.21/
[ "Aaron Traylor", "Roman Feiman", "Ellie Pavlick" ]
A current open question in natural language processing is to what extent language models, which are trained with access only to the form of language, are able to capture the meaning of language. This question is challenging to answer in general, as there is no clear line between meaning and form, but rather meaning con...
2021.acl-short.21
10.18653/v1/2021.acl-short.21
null
null
null
2021.acl-short.22
Enforcing Consistency in Weakly Supervised Semantic Parsing
https://aclanthology.org/2021.acl-short.22/
[ "Nitish Gupta", "Sameer Singh", "Matt Gardner" ]
The predominant challenge in weakly supervised semantic parsing is that of spurious programs that evaluate to correct answers for the wrong reasons. Prior work uses elaborate search strategies to mitigate the prevalence of spurious programs; however, they typically consider only one input at a time. In this work we exp...
2021.acl-short.22
10.18653/v1/2021.acl-short.22
null
2107.05833
title_snapshot
2021.acl-short.23
An Improved Model for Voicing Silent Speech
https://aclanthology.org/2021.acl-short.23/
[ "David Gaddy", "Dan Klein" ]
In this paper, we present an improved model for voicing silent speech, where audio is synthesized from facial electromyography (EMG) signals. To give our model greater flexibility to learn its own input features, we directly use EMG signals as input in the place of hand-designed features used by prior work. Our model u...
2021.acl-short.23
10.18653/v1/2021.acl-short.23
null
2106.01933
title_snapshot
2021.acl-short.24
What’s in the Box? An Analysis of Undesirable Content in the Common Crawl Corpus
https://aclanthology.org/2021.acl-short.24/
[ "Alexandra Luccioni", "Joseph Viviano" ]
Whereas much of the success of the current generation of neural language models has been driven by increasingly large training corpora, relatively little research has been dedicated to analyzing these massive sources of textual data. In this exploratory analysis, we delve deeper into the Common Crawl, a colossal web co...
2021.acl-short.24
10.18653/v1/2021.acl-short.24
null
2105.02732
title_judge
2021.acl-short.25
Continual Quality Estimation with Online Bayesian Meta-Learning
https://aclanthology.org/2021.acl-short.25/
[ "Abiola Obamuyide", "Marina Fomicheva", "Lucia Specia" ]
Most current quality estimation (QE) models for machine translation are trained and evaluated in a static setting where training and test data are assumed to be from a fixed distribution. However, in real-life settings, the test data that a deployed QE model would be exposed to may differ from its training data. In par...
2021.acl-short.25
10.18653/v1/2021.acl-short.25
null
null
null
2021.acl-short.26
A Span-based Dynamic Local Attention Model for Sequential Sentence Classification
https://aclanthology.org/2021.acl-short.26/
[ "Xichen Shang", "Qianli Ma", "Zhenxi Lin", "Jiangyue Yan", "Zipeng Chen" ]
Sequential sentence classification aims to classify each sentence in the document based on the context in which sentences appear. Most existing work addresses this problem using a hierarchical sequence labeling network. However, they ignore considering the latent segment structure of the document, in which contiguous s...
2021.acl-short.26
10.18653/v1/2021.acl-short.26
null
null
null
2021.acl-short.27
How effective is BERT without word ordering? Implications for language understanding and data privacy
https://aclanthology.org/2021.acl-short.27/
[ "Jack Hessel", "Alexandra Schofield" ]
Ordered word sequences contain the rich structures that define language. However, it’s often not clear if or how modern pretrained language models utilize these structures. We show that the token representations and self-attention activations within BERT are surprisingly resilient to shuffling the order of input tokens...
2021.acl-short.27
10.18653/v1/2021.acl-short.27
null
null
null
2021.acl-short.28
WikiSum: Coherent Summarization Dataset for Efficient Human-Evaluation
https://aclanthology.org/2021.acl-short.28/
[ "Nachshon Cohen", "Oren Kalinsky", "Yftah Ziser", "Alessandro Moschitti" ]
Recent works made significant advances on summarization tasks, facilitated by summarization datasets. Several existing datasets have the form of coherent-paragraph summaries. However, these datasets were curated from academic documents that were written for experts, thus making the essential step of assessing the summa...
2021.acl-short.28
10.18653/v1/2021.acl-short.28
null
null
null
2021.acl-short.29
UMIC: An Unreferenced Metric for Image Captioning via Contrastive Learning
https://aclanthology.org/2021.acl-short.29/
[ "Hwanhee Lee", "Seunghyun Yoon", "Franck Dernoncourt", "Trung Bui", "Kyomin Jung" ]
Despite the success of various text generation metrics such as BERTScore, it is still difficult to evaluate the image captions without enough reference captions due to the diversity of the descriptions. In this paper, we introduce a new metric UMIC, an Unreferenced Metric for Image Captioning which does not require ref...
2021.acl-short.29
10.18653/v1/2021.acl-short.29
null
2106.14019
title_snapshot