Agent4Rec

Open Source Research Labs Open Source

Agent4Rec is a recommender system simulator that utilizes 1,000 LLM-empowered generative agents. These agents, initialized from the MovieLens-1M dataset, exhibit diverse social traits and preferences.

Visit Website

Research

Industry

Agent4Rec is a project that showcases the implementation of generative agents in recommendation systems, contributing innovative perspectives for those interested in this technology. This implementation is particularly relevant for the SIGIR 2024 conference, where cutting-edge research in the field will be highlighted. The project's open-source nature allows developers and researchers to explore and adapt the techniques presented.

Agent4Rec Key Features

Allows for customizable simulation settings
Records interaction history of agents
Simulates 1,000 LLM-empowered generative agents
Supports various recommendation algorithms (Random, Pop, MF, MultVAE, LightGCN)
Utilizes the MovieLens-1M dataset

Agent4Rec Use Cases

Exploring user behavior modeling in recommender systems
Research on generative agents in recommendation contexts
Simulating user interactions in recommendation systems
Testing and evaluating recommendation algorithms

Similar AI Agents

GPT Researcher

GPT based autonomous agent that does online comprehensive research on any given topic

Multiagent Debate

An implementation of the paper "Improving Factuality and Reasoning in Language Models through Multiagent Debate"

Teenage AGI

A BabyAGI-inspired agent that can recall infinite memory, "thinks" before making action, and doesn't lose memory after being shutting down