Boosting Meta-Learning for Few-Shot Text Classification via Label-guided Distance Scaling

2026-02-28

AI Summary

This work addresses the challenge in few-shot text classification where randomly selected labeled support samples during testing often provide weak supervision due to poor representativeness, leading to degraded performance. To mitigate this issue, the authors propose Label-guided Distance Scaling (LDS), a novel approach that consistently incorporates label semantics as a supervisory signal during both training and testing. Specifically, LDS employs a label-guided loss to pull each sample closer to its corresponding label representation and introduces a label-guided scaler at test time to dynamically adjust sample embeddings. By integrating meta-learning, label semantic embeddings, and distance metric learning, the method achieves significant improvements over state-of-the-art models across multiple benchmark datasets, demonstrating its effectiveness and robustness.

Technology Category

Machine Learning: Multi-class/Multi-label Learning & Extreme ClassificationNatural Language Processing: Sentence-level Semantics, Textual Inference, etc.Search and Optimization: Learning to Search

Application Category

Economics, Online Markets and Human Computation: Humans versus LLMs for data annotation and labelingSemantics and Knowledge: Methods to enhance, augment, integrate or synergize semantic models such as knowledge graphs and LLMsSearch and Retrieval-Augmented AI: Web learning to rank, online learning, and counterfactual learning for ranking

Abstract

Few-shot text classification aims to recognize unseen classes with limited labeled text samples. Existing approaches focus on boosting meta-learners by developing complex algorithms in the training stage. However, the labeled samples are randomly selected during the testing stage, so they may not provide effective supervision signals, leading to misclassification. To address this issue, we propose a \textbf{L}abel-guided \textbf{D}istance \textbf{S}caling (LDS) strategy. The core of our method is exploiting label semantics as supervision signals in both the training and testing stages. Specifically, in the training stage, we design a label-guided loss to inject label semantic information, pulling closer the sample representations and corresponding label representations. In the testing stage, we propose a Label-guided Scaler which scales sample representations with label semantics to provide additional supervision signals. Thus, even if labeled sample representations are far from class centers, our Label-guided Scaler pulls them closer to their class centers, thereby mitigating the misclassification. We combine two common meta-learners to verify the effectiveness of the method. Extensive experimental results demonstrate that our approach significantly outperforms state-of-the-art models. All datasets and codes are available at https://anonymous.4open.science/r/Label-guided-Text-Classification.

Problem

• Few-shot text classification
• meta-learning
• label semantics
• supervision signals
• misclassification

Innovation

• Label-guided Distance Scaling
• Few-shot Text Classification
• Meta-learning
• Label Semantics
• Distance Scaling