Ensemble Making Few-Shot Learning Stronger

TL;DR

This paper proposes an ensemble approach with fine-tuning and feature attention strategies to enhance few-shot relation learning, effectively reducing variance and outperforming previous models.

Contribution

It introduces a novel ensemble method combined with feature calibration techniques to improve the robustness and accuracy of few-shot relation learning models.

Findings

Significant performance improvement over state-of-the-art models

Effective variance reduction in few-shot relation extraction

Enhanced relation-level feature calibration

Abstract

Few-shot learning has been proposed and rapidly emerging as a viable means for completing various tasks. Many few-shot models have been widely used for relation learning tasks. However, each of these models has a shortage of capturing a certain aspect of semantic features, for example, CNN on long-range dependencies part, Transformer on local features. It is difficult for a single model to adapt to various relation learning, which results in the high variance problem. Ensemble strategy could be competitive on improving the accuracy of few-shot relation extraction and mitigating high variance risks. This paper explores an ensemble approach to reduce the variance and introduces fine-tuning and feature attention strategies to calibrate relation-level features. Results on several few-shot relation learning tasks show that our model significantly outperforms the previous state-of-the-art models.