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

Classic approaches to content moderation typically apply a rule-basedheuristic approach to flag content. While rules are easily customizable andintuitive for humans to interpret, they are inherently fragile and lack theflexibility or robustness needed to moderate the vast amount of undesirablecontent found online today. Recent advances in deep learning have demonstratedthe promise of using highly effective deep neural models to overcome thesechallenges. However, despite the improved performance, these data-driven modelslack transparency and explainability, often leading to mistrust from everydayusers and a lack of adoption by many platforms. In this paper, we present RuleBy Example (RBE): a novel exemplar-based contrastive learning approach forlearning from logical rules for the task of textual content moderation. RBE iscapable of providing rule-grounded predictions, allowing for more explainableand customizable predictions compared to typical deep learning-basedapproaches. We demonstrate that our approach is capable of learning rich ruleembedding representations using only a few data examples. Experimental resultson 3 popular hate speech classification datasets show that RBE is able tooutperform state-of-the-art deep learning classifiers as well as the use ofrules in both supervised and unsupervised settings while providing explainablemodel predictions via rule-grounding.

Rule By Example: Harnessing Logical Rules for Explainable Hate Speech Detection