Review-Incorporated Model-Agnostic Profile Injection Attacks on Recommender Systems

arXiv:2402.09023v1 Announce Type: new
Abstract: Recent studies have shown that recommender systems (RSs) are highly vulnerable to data poisoning attacks. Understanding attack tactics helps improve the robustness of RSs. We intend to develop efficient attack methods that use limited resources to generate high-quality fake user profiles to achieve 1) transferability among black-box RSs 2) and imperceptibility among detectors. In order to achieve these goals, we introduce textual reviews of products to enhance the generation quality of the profiles. Specifically, we …

arxiv attack attacks box cs.ai cs.cr data data poisoning fake high injection injection attacks poisoning poisoning attacks profile profiles quality recommender systems resources review robustness rss studies systems tactics understanding vulnerable