Agent4Rec

Agent4Rec is a recommender system simulator that utilizes 1,000 LLM-empowered generative agents to model diverse social traits and preferences in movie recommendations. Built upon the MovieLens-1M dataset, it explores agent behavior in personalized recommendation environments, offering insights into actions like rating, watching, and interacting with recommendations. The system aids in studying generative agent applications in recommendation tasks and behavioral analysis, simulating the behavior of genuine, independent humans in recommendation environments. It supports various recommendation systems like Random, Pop, MF, MultVAE, and LightGCN, and allows for configurable simulations to explore different settings.