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

The pervasiveness of offensive language on the social network has causedadverse effects on society, such as abusive behavior online. It is urgent todetect offensive language and curb its spread. Existing research shows thatmethods with community structure features effectively improve the performanceof offensive language detection. However, the existing models deal withcommunity structure independently, which seriously affects the effectiveness ofdetection models. In this paper, we propose an end-to-end method based oncommunity structure and text features for offensive language detection(CT-OLD). Specifically, the community structure features are directly capturedby the graph attention network layer, and the text embeddings are taken fromthe last hidden layer of BERT. Attention mechanisms and position encoding areused to fuse these features. Meanwhile, we add user opinion to the communitystructure for representing user features. The user opinion is represented byuser historical behavior information, which outperforms that represented bytext information. Besides the above point, the distribution of users and tweetsis unbalanced in the popular datasets, which limits the generalization abilityof the model. To address this issue, we construct and release a dataset withreasonable user distribution. Our method outperforms baselines with the F1score of 89.94%. The results show that the end-to-end model effectively learnsthe potential information of community structure and text, and user historicalbehavior information is more suitable for user opinion representation.

Detecting Offensive Language on Social Networks: An End-to-end Detection Method based on Graph Attention Networks