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

Despite being able to generate fluent and grammatical text, current Seq2Seqsummarization models still suffering from the unfaithful generation problem. Inthis paper, we study the faithfulness of existing systems from a newperspective of factual robustness which is the ability to correctly generatefactual information over adversarial unfaithful information. We first measure amodel's factual robustness by its success rate to defend against adversarialattacks when generating factual information. The factual robustness analysis ona wide range of current systems shows its good consistency with human judgmentson faithfulness. Inspired by these findings, we propose to improve thefaithfulness of a model by enhancing its factual robustness. Specifically, wepropose a novel training strategy, namely FRSUM, which teaches the model todefend against both explicit adversarial samples and implicit factualadversarial perturbations. Extensive automatic and human evaluation resultsshow that FRSUM consistently improves the faithfulness of various Seq2Seqmodels, such as T5, BART.

FRSUM: Towards Faithful Abstractive Summarization via Enhancing Factual Robustness