Lune

EMNLP Findings2020

Learning to Learn to Disambiguate: Meta-Learning for Few-Shot Word Sense Disambiguation

Nithin Holla, Pushkar Mishra, Helen Yannakoudakis, Ekaterina Shutova

Abstract

The success of deep learning methods hinges on the availability of large training datasets annotated for the task of interest. In contrast to human intelligence, these methods lack versatility and struggle to learn and adapt quickly to new tasks, where labeled data is scarce. Metalearning aims to solve this problem by training a model on a large number of few-shot tasks, with an objective to learn new tasks quickly from a small number of examples. In this paper, we propose a meta-learning framework for few-shot word sense disambiguation (WSD), where the goal is to learn to disambiguate unseen words from only a few labeled instances. Meta-learning approaches have so far been typically tested in an N -way, K-shot classification setting where each task has N classes with K examples per class. Owing to its nature, WSD deviates from this controlled setup and requires the models to handle a large number of highly unbalanced classes. We extend several popular meta-learning approaches to this scenario, and analyze their strengths and weaknesses in this new challenging setting.

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.

11 top-tier papers have 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

  • Self-Supervised Meta-Learning for Few-Shot Natural Language Classification Tasks cites Meta-Learning without Memorization
  • Rare and Zero-shot Word Sense Disambiguation using Z-Reweighting cites Breaking Through the 80% Glass Ceiling: Raising the State of the Art in Word Sense Disambiguation by Incorporating Knowledge Graph Information
  • Improving Meta-learning for Low-resource Text Classification and Generation via Memory Imitation cites Meta-Learning without Memorization
Builds onCited byLeft to right by year; the more cited, the higher.

Cited by top-tier papers11

  1. Improving Meta-learning for Low-resource Text Classification and Generation via Memory ImitationYingxiu Zhao, Zhiliang Tian, Huaxiu Yao, Yinhe Zheng et al.ACL 2022 · 21 citations
  2. Meta-Learning for Fast Cross-Lingual Adaptation in Dependency ParsingAnna Langedijk, Verna Dankers, Phillip Lippe, Sander Bos et al.ACL 2022 · 17 citations
  3. Rare and Zero-shot Word Sense Disambiguation using Z-ReweightingYing Su, Hongming Zhang, Yangqiu Song, Tong ZhangACL 2022 · 12 citations
  4. Self-Supervised Meta-Learning for Few-Shot Natural Language Classification TasksTrapit Bansal, Rishikesh Jha, Tsendsuren Munkhdalai, Andrew McCallumEMNLP 2020 · 9 citations
  5. Predicting emergent linguistic compositions through time: Syntactic frame extension via multimodal chainingLei Yu, Yang XuEMNLP 2021 · 6 citations

Builds on5

  1. Meta-Dataset: A Dataset of Datasets for Learning to Learn from Few ExamplesEleni Triantafillou, Tyler Zhu, Vincent Dumoulin, Pascal Lamblin et al.ICLR 2020 · 692 citations
  2. Meta-Learning without MemorizationMingzhang Yin, George Tucker, Mingyuan Zhou, Sergey Levine et al.ICLR 2020 · 201 citations
  3. Breaking Through the 80% Glass Ceiling: Raising the State of the Art in Word Sense Disambiguation by Incorporating Knowledge Graph InformationMichele Bevilacqua, Roberto NavigliACL 2020 · 145 citations
  4. Meta-learning curiosity algorithmsFerran Alet, Martin F. Schneider, Tomás Lozano-Pérez, Leslie Pack KaelblingICLR 2020 · 67 citations
  5. Tracking by Instance Detection: A Meta-Learning ApproachGuangting Wang, Chong Luo, Xiaoyan Sun, Zhiwei Xiong et al.CVPR 2020

Related papers

  1. Meta-Learning with Variational Semantic Memory for Word Sense DisambiguationYing-Jun Du, Nithin Holla, Xiantong Zhen, Cees Snoek et al.ACL 2021
  2. Meta Learning to Bridge Vision and Language Models for Multimodal Few-Shot LearningIvona Najdenkoska, Xiantong Zhen, Marcel WorringICLR 2023 · 8 citations
  3. Learning to Balance: Bayesian Meta-Learning for Imbalanced and Out-of-distribution TasksHaebeom Lee, Hayeon Lee, Donghyun Na, Saehoon Kim et al.ICLR 2020 · 115 citations
  4. Task-adaptive Label Dependency Transfer for Few-shot Named Entity RecognitionShan Zhang, Bin Cao, Tianming Zhang, Yuqi Liu et al.ACL Findings 2023 · 5 citations
  5. Diverse Distributions of Self-Supervised Tasks for Meta-Learning in NLPTrapit Bansal, Karthick Prasad Gunasekaran, Tong Wang, Tsendsuren Munkhdalai et al.EMNLP 2021 · 27 citations
Dusk over the sea between two cliffs drawn in fine vertical lines