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/

Zendy Plus

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

Zendy Tools

Zendy Open

Distantly-Supervised Named Entity Recognition (DS-NER) is widely used inreal-world scenarios. It can effectively alleviate the burden of annotation bymatching entities in existing knowledge bases with snippets in the text butsuffer from the label noise. Recent works attempt to adopt the teacher-studentframework to gradually refine the training labels and improve the overallrobustness. However, these teacher-student methods achieve limited performancebecause the poor calibration of the teacher network produces incorrectlypseudo-labeled samples, leading to error propagation. Therefore, we propose:(1) Uncertainty-Aware Teacher Learning that leverages the predictionuncertainty to reduce the number of incorrect pseudo labels in theself-training stage; (2) Student-Student Collaborative Learning that allows thetransfer of reliable labels between two student networks instead ofindiscriminately relying on all pseudo labels from its teacher, and furtherenables a full exploration of mislabeled samples rather than simply filteringunreliable pseudo-labeled samples. We evaluate our proposed method on fiveDS-NER datasets, demonstrating that our method is superior to thestate-of-the-art DS-NER methods.

Improving the Robustness of Distantly-Supervised Named Entity  Recognition via Uncertainty-Aware Teacher Learning and Student-Student  Collaborative Learning