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 Tools annual

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 Tools monthly

Global: Zendy Plus monthly

Presents the access of premium content as premium feature

Premium Content

Global: Zendy Plus annual

Global: Zendy Free Plan

Recent pre-trained abstractive summarization systems have started to achievecredible performance, but a major barrier to their use in practice is theirpropensity to output summaries that are not faithful to the input and thatcontain factual errors. While a number of annotated datasets and statisticalmodels for assessing factuality have been explored, there is no clear pictureof what errors are most important to target or where current techniques aresucceeding and failing. We explore both synthetic and human-labeled datasources for training models to identify factual errors in summarization, andstudy factuality at the word-, dependency-, and sentence-level. Ourobservations are threefold. First, exhibited factual errors differsignificantly across datasets, and commonly-used training sets of simplesynthetic errors do not reflect errors made on abstractive datasets like XSum.Second, human-labeled data with fine-grained annotations provides a moreeffective training signal than sentence-level annotations or synthetic data.Finally, we show that our best factuality detection model enables training ofmore factual XSum summarization models by allowing us to identify non-factualtokens in the training data.

Annotating and Modeling Fine-grained Factuality in Summarization