A privacy-preserving federated graph learning framework for threat detection in IoT trigger-action programming

Yongheng Xing, Liang Hu, Xinqi Du, Zhiqi Shen, Juncheng Hu, Feng Wang

Abstract: Highlights•Propose PFTAP framework to detect the threats in TAP rules in IoT.•Propose HieGAN to learn the feature representation of triggers and actions.•Propose to use symmetric encryption and LDP to protect user privacy.•Achieve superior performance in TAP rule threat detection.