EACL 2023

546 papers
The 2023 European Chapter of the Association for Computational Linguistics (EACL) featured 271 papers focusing on social fairness, bias mitigation, and cross-lingual transfer. Research themes included measuring social bias across the gender spectrum and grounding large language models in authoritative clinical evidence. Investigations into model limitations, such as performance variance in few-shot learning and over-reliance on entity names, were also prominent. Prolific contributors included Mohit Bansal, Iryna Gurevych, and Dan Roth. Methodological trends highlighted advancements in zero-shot text classification and multimodal feature fusion. Notable work included the Massive Text Embedding Benchmark (MTEB) and the ConEntail framework for universal zero-shot classification. Other research explored temporal understanding through salient span masking and the use of human gaze feedback to improve model performance. The conference also addressed efficient representations for large-scale vector retrieval and the development of dialogue datasets for retrieving data tables through natural language.

Top Research Directions

Research directions most represented by EACL 2023 papers.

Mitigating Data Scarcity in Low Resource Language Processing 34
The Language Bias of English Centric Benchmarks 23
Zero-Shot Cross-Lingual Transfer in Multilingual Language Models 16
Processing Non-Compositional Semantics in Natural Language 13
The Performance Advantage of Localized Language Pretraining 12
Pre-Training Language Models on Specialized Corpora 11
Factual Consistency in Abstractive Summarization 8
Harmonizing Syntax with Universal Dependencies Treebanks 7
Contextual Reasoning for Implicit Toxicity Detection 7
Expert Validation and Curation of Linguistic Resources 6
Shifting from Static Benchmarks to Dynamic Evaluation 6
The Mechanics of Bias Amplification in Language Models 6
Cross-Lingual Transfer via Linguistic Proximity 6
Back-Translation for Synthetic Data Augmentation 6
Joint Multilingual Training for Cross-Lingual Transfer 6

Top Authors

AuthorPapers
Mohit Bansal6
Heng Ji5
Iryna Gurevych4
Naoaki Okazaki4
Trevor Cohn4
Bhuwan Dhingra3
Chitta Baral3
Dan Roth3
Greg Durrett3
Swaroop Mishra3
Ali Ghodsi2
Anna Korhonen2
Barry Haddow2
Benjamin Van Durme2
Biao Zhang2
Chenghua Lin2
Dongyeop Kang2
Dong Yu2
Edoardo Ponti2
Fangyu Liu2

Impactful Papers

Papers from EACL 2023 that shaped research clusters.

Foundational

Influential

Papers

Title
Realistic Conversational Question Answering with Answer Selection based on Calibrated Confidence and Uncertainty Measurement
Real-Time Visual Feedback to Guide Benchmark Creation: A Human-and-Metric-in-the-Loop Workflow
Reconstructing Language History by Using a Phonological Ontology. An Analysis of German Surnames
RedHOT: A Corpus of Annotated Medical Questions, Experiences, and Claims on Social Media
RedPenNet for Grammatical Error Correction: Outputs to Tokens, Attentions to Spans
Reinforced Sequence Training based Subjective Bias Correction
Representation biases in sentence transformers
Resources and Few-shot Learners for In-context Learning in Slavic Languages
Retrieval-augmented Image Captioning
Retrieval Enhanced Data Augmentation for Question Answering on Privacy Policies
Retrieve-and-Fill for Scenario-based Task-Oriented Semantic Parsing
Revealing Weaknesses of Vietnamese Language Models Through Unanswerable Questions in Machine Reading Comprehension
Revisiting Dependency Length and Intervener Complexity Minimisation on a Parallel Corpus in 35 Languages
RevUp: Revise and Update Information Bottleneck for Event Representation
Robustification of Multilingual Language Models to Real-world Noise in Crosslingual Zero-shot Settings with Robust Contrastive Pretraining
Robustness Challenges in Model Distillation and Pruning for Natural Language Understanding
Romanian Multiword Expression Detection Using Multilingual Adversarial Training and Lateral Inhibition
Salient Span Masking for Temporal Understanding
Scaling Back-Translation with Domain Text Generation for Sign Language Gloss Translation
Scent Mining: Extracting Olfactory Events, Smell Sources and Qualities
Searching for Better Database Queries in the Outputs of Semantic Parsers
Selective In-Context Data Augmentation for Intent Detection using Pointwise V-Information
Self-Adapted Utterance Selection for Suicidal Ideation Detection in Lifeline Conversations
Self-Adaptive Named Entity Recognition by Retrieving Unstructured Knowledge
Self-imitation Learning for Action Generation in Text-based Games
Self-Supervised Unimodal Label Generation Strategy Using Recalibrated Modality Representations for Multimodal Sentiment Analysis
Self-training Reduces Flicker in Retranslation-based Simultaneous Translation
Semantic Frame Induction with Deep Metric Learning
Semantic Parsing for Conversational Question Answering over Knowledge Graphs
Semantic Specialization for Knowledge-based Word Sense Disambiguation
Semi-supervised New Event Type Induction and Description via Contrastive Loss-Enforced Batch Attention
Semi-supervised Relation Extraction via Data Augmentation and Consistency-training
Sentiment as an Ordinal Latent Variable
Shironaam: Bengali News Headline Generation using Auxiliary Information
Shortcomings of Question Answering Based Factuality Frameworks for Error Localization
Shorten the Long Tail for Rare Entity and Event Extraction
Should You Mask 15% in Masked Language Modeling?
SIDLR: Slot and Intent Detection Models for Low-Resource Language Varieties
Silver Data for Coreference Resolution in Ukrainian: Translation, Alignment, and Projection
Simple and Effective Multi-Token Completion from Masked Language Models
Slav-NER: the 4th Cross-lingual Challenge on Recognition, Normalization, Classification, and Linking of Named Entities across Slavic languages
Small-Text: Active Learning for Text Classification in Python
SMATCH++: Standardized and Extended Evaluation of Semantic Graphs
SMHD-GER: A Large-Scale Benchmark Dataset for Automatic Mental Health Detection from Social Media in German
SocBERT: A Pretrained Model for Social Media Text
Social Commonsense for Explanation and Cultural Bias Discovery
Social Influence Dialogue Systems: A Survey of Datasets and Models For Social Influence Tasks
Socratic Question Generation: A Novel Dataset, Models, and Evaluation
SODAPOP: Open-Ended Discovery of Social Biases in Social Commonsense Reasoning Models
Span-based Named Entity Recognition by Generating and Compressing Information

« prev Page 9/11 next »