Bayesian Active Summarization

TL;DR

Bayesian Active Summarization (BAS) integrates active learning with advanced summarization models, enabling effective summarization with limited annotated data and outperforming random selection especially in low-data scenarios.

Contribution

This paper introduces BAS, a novel method combining Bayesian active learning with state-of-the-art summarization models for improved performance on limited data.

Findings

BAS outperforms random selection in small data settings.

BAS enables effective summarization with minimal annotated data.

Large models can be leveraged effectively with BAS for real-world problems.

Abstract

Bayesian Active Learning has had significant impact to various NLP problems, but nevertheless it's application to text summarization has been explored very little. We introduce Bayesian Active Summarization (BAS), as a method of combining active learning methods with state-of-the-art summarization models. Our findings suggest that BAS achieves better and more robust performance, compared to random selection, particularly for small and very small data annotation budgets. Using BAS we showcase it is possible to leverage large summarization models to effectively solve real-world problems with very limited annotated data.