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

Prior work on pretrained sentence embeddings and benchmarks focus on thecapabilities of stand-alone sentences. We propose DiscoEval, a test suite oftasks to evaluate whether sentence representations include broader contextinformation. We also propose a variety of training objectives that makes use ofnatural annotations from Wikipedia to build sentence encoders capable ofmodeling discourse. We benchmark sentence encoders pretrained with our proposedtraining objectives, as well as other popular pretrained sentence encoders onDiscoEval and other sentence evaluation tasks. Empirically, we show that thesetraining objectives help to encode different aspects of information in documentstructures. Moreover, BERT and ELMo demonstrate strong performances overDiscoEval with individual hidden layers showing different characteristics.

Evaluation Benchmarks and Learning Criteria for Discourse-Aware Sentence Representations