ALLWAS: Active Learning on Language models in WASserstein space

TL;DR

ALLWAS introduces a novel active learning method for language models that leverages submodular optimization and Wasserstein barycenters, significantly improving performance on text classification tasks with limited labeled data.

Contribution

The paper proposes a new active learning approach using submodular optimization and Wasserstein barycenters tailored for language models, addressing class imbalance and data scarcity.

Findings

Achieves over 20% relative performance improvement on benchmark datasets.

Effective in scenarios with class imbalance and limited labeled data.

Outperforms existing active learning methods for language models.

Abstract

Active learning has emerged as a standard paradigm in areas with scarcity of labeled training data, such as in the medical domain. Language models have emerged as the prevalent choice of several natural language tasks due to the performance boost offered by these models. However, in several domains, such as medicine, the scarcity of labeled training data is a common issue. Also, these models may not work well in cases where class imbalance is prevalent. Active learning may prove helpful in these cases to boost the performance with a limited label budget. To this end, we propose a novel method using sampling techniques based on submodular optimization and optimal transport for active learning in language models, dubbed ALLWAS. We construct a sampling strategy based on submodular optimization of the designed objective in the gradient domain. Furthermore, to enable learning from few samples, we propose a novel strategy for sampling from the Wasserstein barycenters. Our empirical evaluations on standard benchmark datasets for text classification show that our methods perform significantly better (>20% relative increase in some cases) than existing approaches for active learning on language models.