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 Tools annual

Global: Zendy Free Plan

Global: Zendy Tools monthly

Global: Zendy Plus monthly

Supervised machine learning and deep learning require a large amount oflabeled data, which data scientists obtain in a manual, and time-consumingannotation process. To mitigate this challenge, Active Learning (AL) proposespromising data points to annotators they annotate next instead of a subsequentor random sample. This method is supposed to save annotation effort whilemaintaining model performance. However, practitioners face many AL strategiesfor different tasks and need an empirical basis to choose between them. Surveyscategorize AL strategies into taxonomies without performance indications.Presentations of novel AL strategies compare the performance to a small subsetof strategies. Our contribution addresses the empirical basis by introducing areproducible active learning evaluation (ALE) framework for the comparativeevaluation of AL strategies in NLP. The framework allows the implementation ofAL strategies with low effort and a fair data-driven comparison throughdefining and tracking experiment parameters (e.g., initial dataset size, numberof data points per query step, and the budget). ALE helps practitioners to makemore informed decisions, and researchers can focus on developing new, effectiveAL strategies and deriving best practices for specific use cases. With bestpractices, practitioners can lower their annotation costs. We present a casestudy to illustrate how to use the framework.

ALE: A Simulation-Based Active Learning Evaluation Framework for the Parameter-Driven Comparison of Query Strategies for NLP