AACL 2022

• German
• Italian
• Chinese
• Japanese

TLDRs

• Chasing the Tail with Domain Generalization: A Case Study on Frequency-Enriched Datasets
  • Manoj Kumar, Anna Rumshisky, Rahul Gupta
  • TLDR: We propose an alternative strategy that explicitly uses utterance frequency in training data to learn models that are more robust to unknown distributions.
• Double Trouble: How to not Explain a Text Classifier’s Decisions Using Counterfactuals Synthesized by Masked Language Models?
  • Thang Pham, Trung Bui, Long Mai, Anh Nguyen
  • TLDR: We show that the Deletion-BERT metric used in Input Marginalization is biased towards IM and that making LIME samples more natural via BERT improves LIME accuracy under several ROAR metrics.
• An Empirical Study on Cross-X Transfer for Legal Judgment Prediction
  • Joel Niklaus, Matthias Stürmer, Ilias Chalkidis
  • TLDR: We explore transfer learning techniques on Legal Judgment Prediction using the trilingual Swiss-Judgment-Prediction dataset, including cases written in three languages.
• CNN for Modeling Sanskrit Originated Bengali and Hindi Language
  • Chowdhury Rahman, MD. Hasibur Rahman, Mohammad Rafsan, Mohammed Eunus Ali, Samiha Zakir, Rafsanjani Muhammod
  • TLDR: We propose an end to end trainable memory efficient CNN architecture named CoCNN to handle specific characteristics such as high inflection, morphological richness, flexible word order and phonetical spelling errors of Bengali and Hindi.
• Leveraging Key Information Modeling to Improve Less-Data Constrained News Headline Generation via Duality Fine-Tuning
  • Zhuoxuan Jiang, Lingfeng Qiao, Di Yin, Shanshan Feng, Bo Ren
  • TLDR: We propose a novel duality fine-tuning method for less-data constrained headline generation and show that it can be used to improve language generative models.
• Systematic Evaluation of Predictive Fairness
  • Xudong Han, Aili Shen, Trevor Cohn, Timothy Baldwin, Lea Frermann
  • TLDR: We present a comprehensive evaluation of debiasing methods across multiple tasks, spanning binary classification (Twitter sentiment), multi-class classification (profession prediction), and regression (valence prediction).
• Graph-augmented Learning to Rank for Querying Large-scale Knowledge Graph
  • Hanning Gao, Lingfei Wu, Po Hu, Zhihua Wei, Fangli Xu, Bo Long
  • TLDR: We propose a graph-augmented learning to rank method for knowledge graph question answering based on information retrieval.
• An Embarrassingly Simple Approach for Intellectual Property Rights Protection on Recurrent Neural Networks
  • Zhi Qin Tan, Hao Shan Wong, Chee Seng Chan
  • TLDR: We propose a practical approach for the IPR protection on recurrent neural networks (RNN) without all the bells and whistles of existing IPR solutions.
• WAX: A New Dataset for Word Association eXplanations
  • Chunhua Liu, Trevor Cohn, Simon De Deyne, Lea Frermann
  • TLDR: We present a large, crowd-sourced data set of English word associations with explanations, labeled with high-level relation types, and design several tasks to probe to what extent current pre-trained language models capture the underlying relations.
• Missing Modality meets Meta Sampling (M3S): An Efficient Universal Approach for Multimodal Sentiment Analysis with Missing Modality
  • Haozhe Chi, Minghua Yang, Junhao Zhu, Guanhong Wang, Gaoang Wang
  • TLDR: We propose a simple yet effective meta-sampling approach for multimodal sentiment analysis with missing modalities, namely Missing Modality-based Meta Sampling (M3S).
• SPARQL-to-Text Question Generation for Knowledge-Based Conversational Applications
  • Gwénolé Lecorvé, Morgan Veyret, Quentin Brabant, Lina M. Rojas Barahona
  • TLDR: We present a new set of questions generated by SPARQL queries based on conversational use cases and show that simple questions and frequent templates of queries are usually well processed whereas complex questions and conversational dimensions (coreferences and ellipses) are still difficult to handle.
• S+PAGE: A Speaker and Position-Aware Graph Neural Network Model for Emotion Recognition in Conversation
  • Chen Liang, Jing Xu, Yangkun Lin, Chong Yang, Yongliang Wang
  • TLDR: We propose a novel GNN-based model for emotion recognition in conversation that captures both the speaker and position-aware conversation structure information.
• Grammatical Error Correction Systems for Automated Assessment: Are They Susceptible to Universal Adversarial Attacks?
  • Vyas Raina, Yiting Lu, Mark Gales
  • TLDR: A simple adversarial attack can deceive GEC systems into not correcting grammatical errors.
• This Patient Looks Like That Patient: Prototypical Networks for Interpretable Diagnosis Prediction from Clinical Text
  • Betty van Aken, Jens-Michalis Papaioannou, Marcel Naik, Georgios Eleftheriadis, Wolfgang Nejdl, Felix Gers, Alexander Loeser
  • TLDR: We present ProtoPatient, a novel method based on prototypical networks and label-wise attention for diagnosis prediction from clinical text.
• Cross-lingual Similarity of Multilingual Representations Revisited
  • Maksym Del, Mark Fishel
  • TLDR: We show that the first align, then predict pattern in cross-lingual learning analysis is not only present in masked language models but also in multilingual models with causal language modeling objectives.
• Arabic Dialect Identification with a Few Labeled Examples Using Generative Adversarial Networks
  • Mahmoud Yusuf, Marwan Torki, Nagwa El-Makky
  • TLDR: We extend the transformer-based models, ARBERT and MARBERT, with unlabeled data in a generative adversarial setting using Semi-Supervised Generative Adversarial Networks (SS-GAN).
• Semantic Shift Stability: Efficient Way to Detect Performance Degradation of Word Embeddings and Pre-trained Language Models
  • Shotaro Ishihara, Hiromu Takahashi, Hono Shirai
  • TLDR: We propose a new method to detect time-series performance degradation of word embeddings and pre-trained language models by calculating the degree of semantic shift.
• Neural Text Sanitization with Explicit Measures of Privacy Risk
  • Anthi Papadopoulou, Yunhao Yu, Pierre Lison, Lilja Øvrelid
  • TLDR: We present a novel approach for text sanitization that uses explicit measures of privacy risk to control the trade-off between privacy protection and data utility.
• AGRank: Augmented Graph-based Unsupervised Keyphrase Extraction
  • Haoran Ding, Xiao Luo
  • TLDR: Graph-based unsupervised keyphrase extraction using graph-based deep learning and mutual attention.
• Towards Unified Representations of Knowledge Graph and Expert Rules for Machine Learning and Reasoning
  • Zhepei Wei, Yue Wang, Jinnan Li, Zhining Liu, Erxin Yu, Yuan Tian, Xin Wang, Yi Chang
  • TLDR: We propose a general framework for unifying heterogeneous symbolic knowledge and its representation as a neural architecture for inference and learning.
• Who did what to Whom? Language models and humans respond diversely to features affecting argument hierarchy construction
  • Xiaonan Xu, Haoshuo Chen
  • TLDR: We investigate if transformer-based language models and humans construct argument hierarchy similarly with the effects from telicity, agency, and individuation, using the Chinese structure “NP1+BA/BEI+NP2+VP”.
• CrowdChecked: Detecting Previously Fact-Checked Claims in Social Media
  • Momchil Hardalov, Anton Chernyavskiy, Ivan Koychev, Dmitry Ilvovsky, Preslav Nakov
  • TLDR: We propose a novel algorithm for crowd-based fact-checking of claims in social media for which users have responded with a link to a fact-check article.
• Hate Speech and Offensive Language Detection in Bengali
  • Mithun Das, Somnath Banerjee, Punyajoy Saha, Animesh Mukherjee
  • TLDR: We present a dataset of 10K Bengali posts for hateful and offensive content detection and present several baseline models for such content detection.
• Learning Interpretable Latent Dialogue Actions With Less Supervision
  • Vojtěch Hudeček, Ondřej Dušek
  • TLDR: We present a novel architecture for explainable modeling of task-oriented dialogues with discrete latent variables to represent dialogue actions.
• Named Entity Recognition in Twitter: A Dataset and Analysis on Short-Term Temporal Shifts
  • Asahi Ushio, Francesco Barbieri, Vitor Sousa, Leonardo Neves, Jose Camacho-Collados
  • TLDR: We present a new dataset for Named Entity Recognition in Twitter and analyze the impact of different time periods on language model performance.
• PInKS: Preconditioned Commonsense Inference with Minimal Supervision
  • Ehsan Qasemi, Piyush Khanna, Qiang Ning, Muhao Chen
  • TLDR: We present PInKS, Preconditioned Commonsense Inference with WeaK Supervision, an improved model for reasoning with preconditions through minimum supervision.
• Cross-Lingual Open-Domain Question Answering with Answer Sentence Generation
  • Benjamin Muller, Luca Soldaini, Rik Koncel-Kedziorski, Eric Lind, Alessandro Moschitti
  • TLDR: We present a cross-lingual generative model that produces complete sentences for multilingual and cross-language question answering.
• Discourse Parsing Enhanced by Discourse Dependence Perception
  • Yuqing Xing, Longyin Zhang, Fang Kong, Guodong Zhou
  • TLDR: We propose a new top-down neural model for discourse parsing that learns from discourse dependency and constituency parsing directly to shorten the hierarchical distance of the RST structure.
• Prediction of People’s Emotional Response towards Multi-modal News
  • Ge Gao, Sejin Paik, Carley Reardon, Yanling Zhao, Lei Guo, Prakash Ishwar, Margrit Betke, Derry Tanti Wijaya
  • TLDR: We aim to develop methods for understanding how multimedia news exposure can affect people’s emotional responses, and we especially focus on news content related to gun violence, a very important yet polarizing issue in the U.S.
• AugCSE: Contrastive Sentence Embedding with Diverse Augmentations
  • Zilu Tang, Muhammed Yusuf Kocyigit, Derry Tanti Wijaya
  • TLDR: We present AugCSE, a unified framework to utilize diverse sets of data augmentations to achieve a general-purpose, sentence embedding model.
• Seamlessly Integrating Factual Information and Social Content with Persuasive Dialogue
  • Maximillian Chen, Weiyan Shi, Feifan Yan, Ryan Hou, Jingwen Zhang, Saurav Sahay, Zhou Yu
  • TLDR: We present a novel modular dialogue system framework that seamlessly integrates factual information and social content into persuasive dialogue.
• Dual-Encoder Transformers with Cross-modal Alignment for Multimodal Aspect-based Sentiment Analysis
  • Zhewen Yu, Jin Wang, Liang-Chih Yu, Xuejie Zhang
  • TLDR: We propose a dual-encoder transformer with cross-modal alignment for multimodal aspect-based sentiment analysis.
• AVAST: Attentive Variational State Tracker in a Reinforced Navigator
  • Je-Wei Jang, Mahdin Rohmatillah, Jen-Tzung Chien
  • TLDR: We propose a novel variational approach to approximate belief state distribution for the construction of a reinforced navigator in vision-and-language navigation task.
• Phylogeny-Inspired Adaptation of Multilingual Models to New Languages
  • Fahim Faisal, Antonios Anastasopoulos
  • TLDR: We show how to use language phylogenetic information to improve cross-lingual transfer leveraging closely related languages in a structured, linguistically-informed manner.
• Transferring Knowledge via Neighborhood-Aware Optimal Transport for Low-Resource Hate Speech Detection
  • Tulika Bose, Irina Illina, Dominique Fohr
  • TLDR: We propose a novel training strategy that allows flexible modeling of the relative proximity of neighbors retrieved from a resource-rich corpus to learn the amount of transfer.
• Bag-of-Vectors Autoencoders for Unsupervised Conditional Text Generation
  • Florian Mai, James Henderson
  • TLDR: We propose a novel autoencoder for unsupervised conditional text generation that learns to map an input bag to an output bag, including a novel loss function and neural architecture.
• RecInDial: A Unified Framework for Conversational Recommendation with Pretrained Language Models
  • Lingzhi Wang, Huang Hu, Lei Sha, Can Xu, Daxin Jiang, Kam-Fai Wong
  • TLDR: We propose a novel unified framework for conversational recommendation system that integrates recommendation into the dialog generation by introducing a vocabulary pointer.
• SummVD : An efficient approach for unsupervised topic-based text summarization
  • Gabriel Shenouda, Aurélien Bossard, Oussama Ayoub, Christophe Rodrigues
  • TLDR: We propose a new method for automatic unsupervised extractive summarization of text corpora using singular value decomposition and word clustering.
• Director: Generator-Classifiers For Supervised Language Modeling
  • Kushal Arora, Kurt Shuster, Sainbayar Sukhbaatar, Jason Weston
  • TLDR: We present a unified language modeling architecture that improves generation quality and training performance while avoiding undesirable behaviors.
• VLStereoSet: A Study of Stereotypical Bias in Pre-trained Vision-Language Models
  • Kankan Zhou, Eason Lai, Jing Jiang
  • TLDR: We propose a probing task to measure stereotypical bias in pre-trained vision-language models and propose a new probing task that detects bias by evaluating a model’s tendency to pick stereotypical statements as captions for anti-stereotypical images.
• Dynamic Context Extraction for Citation Classification
  • Suchetha Nambanoor Kunnath, David Pride, Petr Knoth
  • TLDR: We investigate the effect of varying citation context window sizes on model performance in citation intent classification.
• Affective Retrofitted Word Embeddings
  • Sapan Shah, Sreedhar Reddy, Pushpak Bhattacharyya
  • TLDR: We present a novel retrofitting method for updating embeddings of words for their affective meaning.
• Is Encoder-Decoder Redundant for Neural Machine Translation?
  • Yingbo Gao, Christian Herold, Zijian Yang, Hermann Ney
  • TLDR: We propose a new approach for machine translation that uses only the source and target sentences as context.
• SAPGraph: Structure-aware Extractive Summarization for Scientific Papers with Heterogeneous Graph
  • Siya Qi, Lei Li, Yiyang Li, Jin Jiang, Dingxin Hu, Yuze Li, Yingqi Zhu, Yanquan Zhou, Marina Litvak, Natalia Vanetik
  • TLDR: We present a new paper extractive summarization framework based on structure-aware heterogeneous graph for scientific paper summarization.
• Toward Implicit Reference in Dialog: A Survey of Methods and Data
  • Lindsey Vanderlyn, Talita Anthonio, Daniel Ortega, Michael Roth, Ngoc Thang Vu
  • TLDR: We review the state of the art in research regarding the automatic processing of such incompleteness in natural language.
• A Decade of Knowledge Graphs in Natural Language Processing: A Survey
  • Phillip Schneider, Tim Schopf, Juraj Vladika, Mikhail Galkin, Elena Simperl, Florian Matthes
  • TLDR: We present a structured overview of the research landscape in knowledge graphs in NLP and provide a taxonomy of tasks, research types, and contributions.
• Multimodal Generation of Radiology Reports using Knowledge-Grounded Extraction of Entities and Relations
  • Francesco Dalla Serra, William Clackett, Hamish MacKinnon, Chaoyang Wang, Fani Deligianni, Jeff Dalton, Alison Q. O’Neil
  • TLDR: We propose a novel approach for generating text radiology report generation that is state-of-the-art on most of the standard text generation metrics.
• SBERT studies Meaning Representations: Decomposing Sentence Embeddings into Explainable Semantic Features
  • Juri Opitz, Anette Frank
  • TLDR: We learn to induce sentence embeddings that show high correlation to human similarity ratings, but lack interpretability.
• The Lifecycle of “Facts”: A Survey of Social Bias in Knowledge Graphs
  • Angelie Kraft, Ricardo Usbeck
  • TLDR: We investigate factors introducing bias in knowledge graphs and their embedded versions afterward.
• Food Knowledge Representation Learning with Adversarial Substitution
  • Diya Li, Mohammed J Zaki
  • TLDR: We propose a novel approach for generating high quality context-aware recipe and ingredient substitutions over a food knowledge graph.
• Construction Repetition Reduces Information Rate in Dialogue
  • Mario Giulianelli, Arabella Sinclair, Raquel Fernandez
  • TLDR: We propose a new hypothesis that construction usage in dialogue reduces utterance information content, and show that this effect is stronger for repetitions of referential constructions.
• Analogy-Guided Evolutionary Pretraining of Binary Word Embeddings
  • R. Alexander Knipper, Md. Mahadi Hassan, Mehdi Sadi, Shubhra Kanti Karmaker Santu
  • TLDR: We propose a novel genetic algorithm to learn binary word embeddings from scratch, preserving the semantic relationships between words as well as the arithmetic properties of the embeddents themselves.
• Contrastive Video-Language Learning with Fine-grained Frame Sampling
  • Zixu Wang, Yujie Zhong, Yishu Miao, Lin Ma, Lucia Specia
  • TLDR: We propose FineCo (Fine-grained Contrastive Loss for Frame Sampling), an approach to better learn video and language representations with a fine-graining contrastive objective operating on video frames.
• Enhancing Tabular Reasoning with Pattern Exploiting Training
  • Abhilash Shankarampeta, Vivek Gupta, Shuo Zhang
  • TLDR: We present a novel method for improving knowledge and reasoning in tabular inference using pre-trained language models.
• Re-contextualizing Fairness in NLP: The Case of India
  • Shaily Bhatt, Sunipa Dev, Partha Talukdar, Shachi Dave, Vinodkumar Prabhakaran
  • TLDR: We present a framework for fairness evaluation in the context of India and show prediction biases along some of the prominent axes of social disparities in NLP data and models.
• Low-Resource Multilingual and Zero-Shot Multispeaker TTS
  • Florian Lux, Julia Koch, Ngoc Thang Vu
  • TLDR: We show that it is possible to learn to speak a new language using just 5 minutes of training data while retaining the ability to infer the voice of even unseen speakers in the newly learned language.
• Unsupervised Domain Adaptation for Sparse Retrieval by Filling Vocabulary and Word Frequency Gaps
  • Hiroki Iida, Naoaki Okazaki
  • TLDR: We propose an unsupervised domain adaptation method by filling vocabulary and word-frequency gaps.
• KESA: A Knowledge Enhanced Approach To Sentiment Analysis
  • Qinghua Zhao, Shuai Ma, Shuo Ren
  • TLDR: We propose two sentiment-aware auxiliary tasks named sentiment word selection and conditional sentiment prediction and, correspondingly, integrate them into the objective of the downstream task.
• Cross-lingual Few-Shot Learning on Unseen Languages
  • Genta Winata, Shijie Wu, Mayank Kulkarni, Thamar Solorio, Daniel Preotiuc-Pietro
  • TLDR: We explore the problem of cross-lingual transfer in unseen languages, where no unlabeled data is available for pre-training a model.
• Domain-aware Self-supervised Pre-training for Label-Efficient Meme Analysis
  • Shivam Sharma, Mohd Khizir Siddiqui, Md. Shad Akhtar, Tanmoy Chakraborty
  • TLDR: We present two novel multi-modal self-supervised pre-training methods for meme analysis that outperform the fully supervised baseline on all three tasks of the Memotion challenge.
• A Prompt Array Keeps the Bias Away: Debiasing Vision-Language Models with Adversarial Learning
  • Hugo Berg, Siobhan Hall, Yash Bhalgat, Hannah Kirk, Aleksandar Shtedritski, Max Bain
  • TLDR: We investigate and mitigate various bias measures and apply ranking metrics for image-text representations.
• Some Languages are More Equal than Others: Probing Deeper into the Linguistic Disparity in the NLP World
  • Surangika Ranathunga, Nisansa de Silva
  • TLDR: We provide a comprehensive analysis of the linguistic disparity in the languages of the world and provide possible reasons for this disparity.
• Neural Readability Pairwise Ranking for Sentences in Italian Administrative Language
  • Martina Miliani, Serena Auriemma, Fernando Alva-Manchego, Alessandro Lenci
  • TLDR: We investigate the behavior of a Neural Pairwise Ranking Model for sentence-level readability assessment of Italian administrative texts.
• Delivering Fairness in Human Resources AI: Mutual Information to the Rescue
  • Leo Hemamou, William Coleman
  • TLDR: We propose to minimize the MI between a candidate’s name and a latent representation of their CV or short biography to mitigate bias from sensitive variables without requiring the collection of these variables.
• Not another Negation Benchmark: The NaN-NLI Test Suite for Sub-clausal Negation
  • Thinh Hung Truong, Yulia Otmakhova, Timothy Baldwin, Trevor Cohn, Jey Han Lau, Karin Verspoor
  • TLDR: We present a new test suite for natural language inference that captures sub-clausal negation and quantification in natural language models.
• HaRiM^+: Evaluating Summary Quality with Hallucination Risk
  • Seonil (Simon) Son, Junsoo Park, Jeong-in Hwang, Junghwa Lee, Hyungjong Noh, Yeonsoo Lee
  • TLDR: We propose a new metric for summarization that measures hallucination risk and a new way to measure the quality of generated summaries.
• The lack of theory is painful: Modeling Harshness in Peer Review Comments
  • Rajeev Verma, Rajarshi Roychoudhury, Tirthankar Ghosal
  • TLDR: We present a dataset of peer-review comments with their real-valued harshness scores and show that this dataset can help to make review reports less hurtful and more welcoming.
• Dual Mechanism Priming Effects in Hindi Word Order
  • Sidharth Ranjan, Marten van Schijndel, Sumeet Agarwal, Rajakrishnan Rajkumar
  • TLDR: We show that different priming influences are separable from one another, and show that multiple different cognitive mechanisms underlie priming.
• Unsupervised Single Document Abstractive Summarization using Semantic Units
  • Jhen-Yi Wu, Ying-Jia Lin, Hung-Yu Kao
  • TLDR: We propose a novel unsupervised model for abstractive summarization that learns the frequency of each semantic unit in the source text and uses it to generate summaries from the filtered sentences.
• Detecting Incongruent News Articles Using Multi-head Attention Dual Summarization
  • Sujit Kumar, Gaurav Kumar, Sanasam Ranbir Singh
  • TLDR: We propose a novel multi-head attention dual summary based method for detecting incongruent news articles.
• Meta-Learning based Deferred Optimisation for Sentiment and Emotion aware Multi-modal Dialogue Act Classification
  • Tulika Saha, Aditya Prakash Patra, Sriparna Saha, Pushpak Bhattacharyya
  • TLDR: Dialogue Act Classification that determines the communicative intention of an utterance has been investigated widely over the years as a standalone task. But the emotional state of the speaker has a considerable effect on its pragmatic content.
• Enhancing Financial Table and Text Question Answering with Tabular Graph and Numerical Reasoning
  • Rungsiman Nararatwong, Natthawut Kertkeidkachorn, Ryutaro Ichise
  • TLDR: We propose a graph neural network-based evidence extraction module for tables and numerical reasoning for financial question answering.
• Fine-grained Contrastive Learning for Definition Generation
  • Hengyuan Zhang, Dawei Li, Shiping Yang, Yanran Li
  • TLDR: We propose a novel contrastive learning method for definition generation, which can generate more detailed semantic representations from the definition sequence encoding.
• Hengam: An Adversarially Trained Transformer for Persian Temporal Tagging
  • Sajad Mirzababaei, Amir Hossein Kargaran, Hinrich Schütze, Ehsaneddin Asgari
  • TLDR: We present an adversarially trained transformer for Persian temporal tagging outperforming state-of-the-art approaches on a diverse and manually created dataset.
• What’s Different between Visual Question Answering for Machine “Understanding” Versus for Accessibility?
  • Yang Trista Cao, Kyle Seelman, Kyungjun Lee, Hal Daumé III
  • TLDR: We evaluate discrepancies between machine “understanding” and accessibility datasets for visual question answering and show how to improve them.
• Persona or Context? Towards Building Context adaptive Personalized Persuasive Virtual Sales Assistant
  • Abhisek Tiwari, Sriparna Saha, Shubhashis Sengupta, Anutosh Maitra, Roshni Ramnani, Pushpak Bhattacharyya
  • TLDR: We propose a novel end-to-end multi-modal persuasive dialogue system incorporated with a personalized persuasive module aided goal controller and goal persuader.
• Legal Case Document Summarization: Extractive and Abstractive Methods and their Evaluation
  • Abhay Shukla, Paheli Bhattacharya, Soham Poddar, Rajdeep Mukherjee, Kripabandhu Ghosh, Pawan Goyal, Saptarshi Ghosh
  • TLDR: We present several different extractive and abstractive summarization methods for legal case document summarization and evaluate their performance.
• FPC: Fine-tuning with Prompt Curriculum for Relation Extraction
  • Sicheng Yang, Dandan Song
  • TLDR: We propose a novel method Fine-tuning with Prompt Curriculum for relation extraction using auxiliary prompt-based fine-tunings and a prompt learning curriculum.
• Dead or Murdered? Predicting Responsibility Perception in Femicide News Reports
  • Gosse Minnema, Sara Gemelli, Chiara Zanchi, Tommaso Caselli, Malvina Nissim
  • TLDR: We show that different linguistic expressions of gender-based violence can trigger different perceptions of responsibility, and that such perceptions can be modelled automatically.
• PESE: Event Structure Extraction using Pointer Network based Encoder-Decoder Architecture
  • Alapan Kuila, Sudeshna Sarkar
  • TLDR: We present an end-to-end solution to the task of event extraction by exploiting the interactions among event participants and presenting a truly end- to-end model for event extraction.
• How do we get there? Evaluating transformer neural networks as cognitive models for English past tense inflection
  • Xiaomeng Ma, Lingyu Gao
  • TLDR: We train a set of transformer models with different settings to examine their behavior on the quasi-regularity of the verbs in English.
• Characterizing and addressing the issue of oversmoothing in neural autoregressive sequence modeling
  • Ilia Kulikov, Maksim Eremeev, Kyunghyun Cho
  • TLDR: We propose to explicitly minimize the oversmoothing rate of neural machine translation models and show that this is the main reason behind the degenerate case of overly probable short sequences in a neural autoregressive model.
• Identifying Weaknesses in Machine Translation Metrics Through Minimum Bayes Risk Decoding: A Case Study for COMET
  • Chantal Amrhein, Rico Sennrich
  • TLDR: We show that neural metrics in machine translation systems are not sensitive enough to discrepancies in numbers and named entities.
• Whodunit? Learning to Contrast for Authorship Attribution
  • Bo Ai, Yuchen Wang, Yugin Tan, Samson Tan
  • TLDR: We propose to learn author-specific language representations by fine-tuning pre-trained generic language representations with a contrastive objective (Contra-X).
• Higher-Order Dependency Parsing for Arc-Polynomial Score Functions via Gradient-Based Methods and Genetic Algorithm
  • Xudong Zhang, Joseph Le Roux, Thierry Charnois
  • TLDR: We present a novel method for higher-order dependency parsing which takes advantage of the general form of score functions written as arc-polynomials, a general framework which encompasses common higher- order score functions, and includes new ones.
• Underspecification in Scene Description-to-Depiction Tasks
  • Ben Hutchinson, Jason Baldridge, Vinodkumar Prabhakaran
  • TLDR: We present a conceptual framework for addressing the ambiguity and underpredictability of multimodal image+text systems, and propose strategies for addressing these challenges.
• COFAR: Commonsense and Factual Reasoning in Image Search
  • Prajwal Gatti, Abhirama Subramanyam Penamakuri, Revant Teotia, Anand Mishra, Shubhashis Sengupta, Roshni Ramnani
  • TLDR: We present a unified framework for commonsense and factual reasoning in image search that leverages visual content and grounded knowledge to learn alignment between images and search queries.