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

Sentence semantic matching is one of the fundamental tasks in naturallanguage processing, which requires an agent to determine the semantic relationamong input sentences. Recently, deep neural networks have achieved impressiveperformance in this area, especially BERT. Despite the effectiveness of thesemodels, most of them treat output labels as meaningless one-hot vectors,underestimating the semantic information and guidance of relations that theselabels reveal, especially for tasks with a small number of labels. To addressthis problem, we propose a Relation of Relation Learning Network (R2-Net) forsentence semantic matching. Specifically, we first employ BERT to encode theinput sentences from a global perspective. Then a CNN-based encoder is designedto capture keywords and phrase information from a local perspective. To fullyleverage labels for better relation information extraction, we introduce aself-supervised relation of relation classification task for guiding R2-Net toconsider more about labels. Meanwhile, a triplet loss is employed todistinguish the intra-class and inter-class relations in a finer granularity.Empirical experiments on two sentence semantic matching tasks demonstrate thesuperiority of our proposed model. As a byproduct, we have released the codesto facilitate other researches.

R$^2$-Net: Relation of Relation Learning Network for Sentence Semantic Matching