(NeurIPS' 24) Large Language Models as Urban Residents: An LLM Agent Framework for Personal Mobility Generation
Welcome to the official implementation of LLMob, as described in our paper Large Language Models as Urban Residents: An LLM Agent Framework for Personal Mobility Generation. This project demonstrates how Large Language Models (LLMs) can be leveraged to generate personal mobility trajectories based on real-world data.
LLMob is an intuitive framework that builds reasoning logic for LLMs in the context of personal activity trajectory generation.
Figure 1: The LLMob Framework Architecture.
Figure 2: Illustration of activity trajectory generated by LLM agent.
- ./simulator/engine/person.py: Generate personal activity trajectory according to real-world check-in data.
- ./simulator/engine/functions/traj_infer.py: Personal activity trajectory generation function.
- ./simulator/engine/functions/PISC.py: Personal activity pattern identification function.
- ./simulator/engine/memory/retrieval_helper.py: Function related to motivation retrieval.
- ./simulator/prompt_template: Prompt template used in this project.
To get started with LLMob, follow these steps:
git clone https://github.com/Wangjw6/LLMob.git
cd LLMob
conda env create -f environment.yml
conda activate llm
python run_anonymized.py You should also add your own OpenAI API key in the ./config/key.yaml file.
If you would like to cite our work, please use:
@article{wang2024large,
title={Large language models as urban residents: An llm agent framework for personal mobility generation},
author={Wang, Jiawei and Jiang, Renhe and Yang, Chuang and Wu, Zengqing and Onizuka, Makoto and Shibasaki, Ryosuke and Koshizuka, Noboru and Xiao, Chuan},
journal={arXiv preprint arXiv:2402.14744},
year={2024}
}
Our implementation adapts several open-source ChatGPT application and have extensively modified it to our purposes. We thank the authors for sharing their implementations and related resources:
The raw data used in this project is from Foursquare API. We select the data with enough records and preprocess them before using in our project.