Lune

ACL Findings2025

Causal Denoising Prototypical Network for Few-Shot Multi-label Aspect Category Detection

Jin Cui, Xinfeng Wang, Yoshimi Suzuki, Fumiyo Fukumoto

Abstract

The multi-label aspect category detection (MACD) task has attracted great attention in sentiment analysis. Many recent methods have formulated the MACD task by learning robust prototypes to represent categories with limited support samples. However, few of them address the noise categories in the support set that hinder their models from effective prototype generations. To this end, we propose a causal denoising prototypical network (CDPN) for few-shot MACD. We reveal the underlying relation between causal inference and contrastive learning, and present causal contrastive learning (CCL) using discrete and continuous noise as negative samples. We empirically found that CCL can (1) prevent models from overly predicting more categories and (2) mitigate semantic ambiguity issues among categories. Experimental results show that CDPN outperforms competitive baselines. Our code is available online: https://github.com/cuijin-23/CPDN.

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.

Questions to start from

Lune can draw on

the full paper · related work

Get the answer

Free to start. No credit card required.

Lineage

Builds onCited byLeft to right by year; the more cited, the higher.

Builds on29

  1. Representation Learning via Invariant Causal MechanismsJovana Mitrovic, Brian McWilliams, Jacob C. Walker, Lars Holger Buesing et al.ICLR 2021 · 281 citations
  2. Clicks can be Cheating: Counterfactual Recommendation for Mitigating Clickbait IssueWenjie Wang, Fuli Feng, Xiangnan He, Hanwang Zhang et al.SIGIR 2021 · 173 citations
  3. CauseRec: Counterfactual User Sequence Synthesis for Sequential RecommendationShengyu Zhang, Dong Yao, Zhou Zhao, Tat-Seng Chua et al.SIGIR 2021 · 118 citations
  4. JointCL: A Joint Contrastive Learning Framework for Zero-Shot Stance DetectionBin Liang, Qinglin Zhu, Xiang Li, Min Yang et al.ACL 2022 · 117 citations
  5. Robustness to Spurious Correlations in Text Classification via Automatically Generated CounterfactualsZhao Wang, Aron CulottaAAAI 2021 · 114 citations

Related papers

  1. Label-Driven Denoising Framework for Multi-Label Few-Shot Aspect Category DetectionFei Zhao, Yuchen Shen, Zhen Wu, Xinyu DaiEMNLP Findings 2022 · 11 citations
  2. Multi-Label Few-Shot Learning for Aspect Category DetectionMengting Hu, Shiwan Zhao, Honglei Guo, Chao Xue et al.ACL 2021
  3. Label-enhanced Prototypical Network with Contrastive Learning for Multi-label Few-shot Aspect Category DetectionHan Liu, Feng Zhang, Xiaotong Zhang, Siyang Zhao et al.KDD 2022 · 23 citations
  4. Variational Hybrid-Attention Framework for Multi-Label Few-Shot Aspect Category DetectionCheng Peng, Ke Chen, Lidan Shou, Gang ChenAAAI 2024 · 3 citations
  5. Zero-Shot Conversational Stance Detection: Dataset and ApproachesYuzhe Ding, Kang He, Bobo Li, Li Zheng et al.ACL Findings 2025 · 4 citations
Dusk over the sea between two cliffs drawn in fine vertical lines