English (United Kingdom)

https://curated-unify.zendy.io/wp-json/zendy-region/v1/featured_content/oa?rat=en

https://curated-unify.zendy.io/wp-json/zendy-region/v1/highlighted_journal/

Global: Zendy Plus annual

Presents the access of premium content as premium feature

Premium Content

Presents the keyphrase highlighting as premium feature

Keyphrase Highlighting

Presents the summarisation as premium feature

Summarisation

Insights

Presents the pdf analysis as premium feature

PDF Analysis

Presents the zaia usage as premium feature

ZAIA

Global: Zendy Plus monthly

Global: Zendy Tools annual

Global: Zendy Tools monthly

Global: Zendy Free Plan

Models pretrained with self-supervised objectives on large text corporaachieve state-of-the-art performance on English text summarization tasks.However, these models are typically fine-tuned on hundreds of thousands of datapoints, an infeasible requirement when applying summarization to new, nichedomains. In this work, we introduce a novel and generalizable method, calledWikiTransfer, for fine-tuning pretrained models for summarization in anunsupervised, dataset-specific manner. WikiTransfer fine-tunes pretrainedmodels on pseudo-summaries, produced from generic Wikipedia data, which containcharacteristics of the target dataset, such as the length and level ofabstraction of the desired summaries. WikiTransfer models achievestate-of-the-art, zero-shot abstractive summarization performance on theCNN-DailyMail dataset and demonstrate the effectiveness of our approach onthree additional diverse datasets. These models are more robust to noisy dataand also achieve better or comparable few-shot performance using 10 and 100training examples when compared to few-shot transfer from other summarizationdatasets. To further boost performance, we employ data augmentation viaround-trip translation as well as introduce a regularization term for improvedfew-shot transfer. To understand the role of dataset aspects in transferperformance and the quality of the resulting output summaries, we further studythe effect of the components of our unsupervised fine-tuning data and analyzefew-shot performance using both automatic and human evaluation.

Improving Zero and Few-Shot Abstractive Summarization with Intermediate Fine-tuning and Data Augmentation