Paper

Overview

Agent4Rec is a user simulator for recommender systems that uses LLM-empowered generative agents. These agents have profile, memory, and action modules to simulate realistic user behavior. It aims to bridge the gap between offline metrics and online performance in recommendation evaluation.

Best for

Best for
Researchers and developers building or evaluating recommender systems who need a realistic user simulation

Use cases

• Simulating user interactions to evaluate recommender system algorithms
• Testing recommendation strategies in a controlled, reproducible environment
• Studying user behavior patterns with LLM-based agents

Notes

Agent4Rec is a user simulator for recommender systems that uses LLM-empowered generative agents. These agents have profile, memory, and action modules to simulate realistic user behavior. It aims to bridge the gap between offline metrics and online performance in recommendation evaluation.

Use cases

• Simulating user interactions to evaluate recommender system algorithms
• Testing recommendation strategies in a controlled, reproducible environment
• Studying user behavior patterns with LLM-based agents

Pros

• Leverages LLMs to produce more realistic and diverse user behaviors
• Modular design allows customization of agent profiles and memory
• Addresses the known disconnect between offline metrics and online performance

Cons

• Research prototype not yet production-ready
• May not fully capture the complexity of real human users
• Requires significant computational resources to run LLM agents

Indexed from awesome-ai-agents and enriched against its public facts.