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

Social platforms such as Twitter are under siege from a multitude offraudulent users. In response, social bot detection tasks have been developedto identify such fake users. Due to the structure of social networks, themajority of methods are based on the graph neural network(GNN), which issusceptible to attacks. In this study, we propose a node injection-basedadversarial attack method designed to deceive bot detection models. Notably,neither the target bot nor the newly injected bot can be detected when a newbot is added around the target bot. This attack operates in a black-boxfashion, implying that any information related to the victim model remainsunknown. To our knowledge, this is the first study exploring the resilience ofbot detection through graph node injection. Furthermore, we develop anattribute recovery module to revert the injected node embedding from the graphembedding space back to the original feature space, enabling the adversary tomanipulate node perturbation effectively. We conduct adversarial attacks onfour commonly used GNN structures for bot detection on two widely useddatasets: Cresci-2015 and TwiBot-22. The attack success rate is over 73\% andthe rate of newly injected nodes being detected as bots is below 13\% on thesetwo datasets.

My Brother Helps Me: Node Injection Based Adversarial Attack on Social Bot Detection