NAACL2022 对比学习论文

目录

Long papers

short papers

Industry Track

---

Long papers

[1] HiURE: Hierarchical Exemplar Contrastive Learning for Unsupervised Relation Extraction

[2] CONFIT: Toward Faithful Dialogue Summarization with Linguistically-Informed Contrastive Fine-tuning

[3] Cross-modal Contrative Learning for Speech Translation

[4] DiffCSE: Difference-based Contrastive Learning for Sentence Embeddings

[5] Revisit Overconfidence for OOD Detection: Reassigned Contrastive Learning with Adaptive Class-dependent Threshold

[6] EASE: Entity-Aware Contrastive Learning of Sentence Embedding

[7] Contrastive Representation Learning for Cross-Document Coreference Resolution of Events and Entities

[8] Domain Confused Contrastive Learning for Unsupervised Domain Adaptation

[9] Intent Detection and Discovery from User Logs via Deep Semi-Supervised Contrastive Clustering

[10] Label Anchored Contrastive Learning for Language Understanding

short papers

[11] MCSE: Multimodal Contrastive Learning of Sentence Embeddings

[12] Contrastive Learning for Prompt-based Few-shot Language Learners

[13] Label Refinement via Contrastive Learning for Distantly-Supervised Named Entity Recognition

[14] Zero-Shot Event Detection Based on Ordered Contrastive Learning and Prompt-Based Prediction

[15] Detect Rumors in Microblog Posts for Low-Resource Domains via Adversarial Contrastive Learning

[16] TaCL: Improving BERT Pre-training with Token-aware Contrastive Learning

[17] RCL: Relation Contrastive Learning for Zero-Shot Relation Extraction

[18] CLMLF:A Contrastive Learning and Multi-Layer Fusion Method for Multimodal Sentiment Detection

[19] To Answer or Not To Answer? Improving Machine Reading Comprehension Model with Span-based Contrastive Learning

[20] CODE-MVP: Learning to Represent Source Code from Multiple Views with Contrastive Pre-Training

[21] Self-Supervised Contrastive Learning with Adversarial Perturbations for Defending Word Substitution-based Attacks

Industry Track

[22] CREATER: CTR-driven Advertising Text Generation with Controlled Pre-Training and Contrastive Fine-Tuning