SemanticShield: LLM-Powered Audits Expose Shilling Attacks in Recommender Systems

Open in new window