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

We present a new approach to evaluate computational models for the task oftext understanding by the means of out-of-context error detection. Through thenovel design of our automated modification process, existing large-scale datasources can be adopted for a vast number of text understanding tasks. The datais thereby altered on a semantic level, allowing models to be tested against achallenging set of modified text passages that require to comprise a broadernarrative discourse. Our newly introduced task targets actual real-worldproblems of transcription and translation systems by inserting authenticout-of-context errors. The automated modification process is applied to the2016 TEDTalk corpus. Entirely automating the process allows the adoption ofcomplete datasets at low cost, facilitating supervised learning procedures anddeeper networks to be trained and tested. To evaluate the quality of themodification algorithm a language model and a supervised binary classificationmodel are trained and tested on the altered dataset. A human baselineevaluation is examined to compare the results with human performance. Theoutcome of the evaluation task indicates the difficulty to detect semanticerrors for machine-learning algorithms and humans, showing that the errorscannot be identified when limited to a single sentence.

Automated Evaluation of Out-of-Context Errors