Efficient Few-Shot Fine-Tuning for Opinion Summarization

TL;DR

This paper introduces an efficient few-shot fine-tuning approach for opinion summarization using adapters, which improves summary quality and coherence on small datasets by leveraging self-supervised pre-training on unannotated reviews.

Contribution

The paper proposes a novel adapter-based fine-tuning method with self-supervised pre-training on unannotated reviews, enhancing opinion summarization performance in low-resource settings.

Findings

Adapter pre-training improves ROUGE-L scores by 2.0 and 1.3 points on Amazon and Yelp datasets.

Self-supervised pre-training reduces disfluencies and semantic errors in summaries.

Query-based adapter pre-training enhances summary coherence and reduces redundancies.

Abstract

Abstractive summarization models are typically pre-trained on large amounts of generic texts, then fine-tuned on tens or hundreds of thousands of annotated samples. However, in opinion summarization, large annotated datasets of reviews paired with reference summaries are not available and would be expensive to create. This calls for fine-tuning methods robust to overfitting on small datasets. In addition, generically pre-trained models are often not accustomed to the specifics of customer reviews and, after fine-tuning, yield summaries with disfluencies and semantic mistakes. To address these problems, we utilize an efficient few-shot method based on adapters which, as we show, can easily store in-domain knowledge. Instead of fine-tuning the entire model, we add adapters and pre-train them in a task-specific way on a large corpus of unannotated customer reviews, using held-out reviews as pseudo summaries. Then, fine-tune the adapters on the small available human-annotated dataset. We show that this self-supervised adapter pre-training improves summary quality over standard fine-tuning by 2.0 and 1.3 ROUGE-L points on the Amazon and Yelp datasets, respectively. Finally, for summary personalization, we condition on aspect keyword queries, automatically created from generic datasets. In the same vein, we pre-train the adapters in a query-based manner on customer reviews and then fine-tune them on annotated datasets. This results in better-organized summary content reflected in improved coherence and fewer redundancies.