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

ive summarization models often generate inconsistent summariescontaining factual errors or hallucinated content. Recent works focus oncorrecting factual errors in generated summaries via post-editing. Suchcorrection models are trained using adversarial non-factual summariesconstructed using heuristic rules for injecting errors. However, generatingnon-factual summaries using heuristics often does not generalize well to actualmodel errors. In this work, we propose to generate hard, representativesynthetic examples of non-factual summaries through infilling language models.With this data, we train a more robust fact-correction model to post-edit thesummaries to improve factual consistency. Through quantitative and qualitativeexperiments on two popular summarization datasets -- CNN/DM and XSum -- we showthat our approach vastly outperforms prior methods in correcting erroneoussummaries. Our model -- FactEdit -- improves factuality scores by over ~11points on CNN/DM and over ~31 points on XSum on average across multiplesummarization models, producing more factual summaries while maintainingcompetitive summarization quality.

Correcting Diverse Factual Errors in Abstractive Summarization via Post-Editing and Language Model Infilling