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

We investigate the problem of determining the predictive confidence (or,conversely, uncertainty) of a neural classifier through the lens oflow-resource languages. By training models on sub-sampled datasets in threedifferent languages, we assess the quality of estimates from a wide array ofapproaches and their dependence on the amount of available data. We find thatwhile approaches based on pre-trained models and ensembles achieve the bestresults overall, the quality of uncertainty estimates can surprisingly sufferwith more data. We also perform a qualitative analysis of uncertainties onsequences, discovering that a model's total uncertainty seems to be influencedto a large degree by its data uncertainty, not model uncertainty. All modelimplementations are open-sourced in a software package.

Exploring Predictive Uncertainty and Calibration in NLP: A Study on the Impact of Method & Data Scarcity