Lune

EMNLP Findings2024

Enhancing Multi-Label Text Classification under Label-Dependent Noise: A Label-Specific Denoising Framework

Pengyu Xu, Liping Jing, Jian Yu

Abstract

Recent advancements in noisy multi-label text classification have primarily relied on the classconditional noise (CCN) assumption, which treats each label independently undergoing label flipping to generate noisy labels. However, in real-world scenarios, noisy labels often exhibit dependencies with true labels.

In this study, we validate through hypothesis testing that real-world datasets are unlikely to adhere to the CCN assumption, indicating that label noise is dependent on the current labels. To address this, we introduce a label-specific denoising framework designed to counteract label-dependent noise. The framework initially presents a holistic selection metric that evaluates noisy labels by concurrently considering loss information, ranking information, and feature centroid. Subsequently, it identifies and corrects noisy labels individually for each label category in a fine-grained manner. Extensive experiments on benchmark datasets demonstrate the effectiveness of our method under both synthetic and real-world noise conditions, significantly improving performance over existing state-of-the-art models.

Ask Lune

Know this paper in minutes, not an afternoon.

Lune can read every page of it and the top-tier papers that cite it, and gives you its claims, its evidence and its limits, each one cited to the paper it comes from. No skimming, no AI slop.

1 top-tier paper has cited it since. Know what they found before your reviewers do.

Questions to start from

Lune can draw on

the full paper · the top-tier papers that cite it · related work

Get the answer

Free to start. No credit card required.

Lineage

  • Beyond Class-Conditional Assumption: A Primary Attempt to Combat Instance-Dependent Label Noise cites DivideMix: Learning with Noisy Labels as Semi-supervised Learning
  • Selective-Supervised Contrastive Learning with Noisy Labels cites DivideMix: Learning with Noisy Labels as Semi-supervised Learning
Builds onCited byLeft to right by year; the more cited, the higher.

Cited by top-tier papers1

  1. Prompt-Induced Linguistic Fingerprints for LLM-Generated Fake News DetectionChi Wang, Min Gao, Zongwei Wang, Junwei Yin et al.WWW 2026 · 3 citations

Builds on17

  1. DivideMix: Learning with Noisy Labels as Semi-supervised LearningJunnan Li, Richard Socher, Steven C. H. HoiICLR 2020 · 1,326 citations
  2. Selective-Supervised Contrastive Learning with Noisy LabelsShikun Li, Xiaobo Xia, Shiming Ge, Tongliang LiuCVPR 2022 · 201 citations
  3. Partial Multi-Label Learning with Noisy Label IdentificationMing-Kun Xie, Sheng-Jun HuangAAAI 2020 · 179 citations
  4. Beyond Class-Conditional Assumption: A Primary Attempt to Combat Instance-Dependent Label NoisePengfei Chen, Junjie Ye, Guangyong Chen, Jingwei Zhao et al.AAAI 2021 · 156 citations
  5. Does Head Label Help for Long-Tailed Multi-Label Text ClassificationLin Xiao, Xiangliang Zhang, Liping Jing, Chi Huang et al.AAAI 2021 · 73 citations

Related papers

  1. Learning with Feature-Dependent Label Noise: A Progressive ApproachYikai Zhang, Songzhu Zheng, Pengxiang Wu, Mayank Goswami et al.ICLR 2021 · 184 citations
  2. Noise-Robust Learning from Multiple Unsupervised Sources of Inferred LabelsAmila Silva, Ling Luo, Shanika Karunasekera, Christopher LeckieAAAI 2022 · 11 citations
  3. LANE: Label-Aware Noise Elimination for Fine-Grained Text ClassificationTiberiu Sosea, Cornelia CarageaICLR 2026
  4. Holistic Label Correction for Noisy Multi-Label ClassificationXiaobo Xia, Jiankang Deng, Wei Bao, Yuxuan Du et al.ICCV 2023 · 13 citations
  5. Adaptive Textual Label Noise Learning based on Pre-trained ModelsShaohuan Cheng, Wenyu Chen, Mingsheng Fu, Xuanting Xie et al.EMNLP Findings 2023 · 1 citation
Dusk over the sea between two cliffs drawn in fine vertical lines