This repo is the source code of the paper "Diet-ODIN: A Novel Framework for Opioid Misuse Detection with Interpretable Dietary Patterns" Check our paper here, and check our proof-of-concept system demo here.
- python==3.8.18
- pytorch==2.1.0
- torch-geometric==2.4.0
To install all requirements for the project using conda:
conda env create -f environment.yml
Please download the benchmark graph dataset from this link, and put the graph under the directory of processed_data/.
This is unnecessary for the reproduction. But the link also contains the structure and supporting files for the graph construction. To download the raw data, please visit the offical NHANES dataset site here. The data should be put within the structure provided. And you can reproduce the graph executing notebooks under code/preprocessing/.
To reproduce the result, please excute the following command under the code/ directory.
python main.py
@inproceedings{zhang2024diet,
title={Diet-ODIN: A Novel Framework for Opioid Misuse Detection with Interpretable Dietary Patterns},
author={Zhang, Zheyuan and Wang, Zehong and Hou, Shifu and Hall, Evan and Bachman, Landon and White, Jasmine and Galassi, Vincent and Chawla, Nitesh V and Zhang, Chuxu and Ye, Yanfang},
booktitle={Proceedings of the 30th ACM SIGKDD Conference on Knowledge Discovery and Data Mining},
pages={6312--6323},
year={2024}
}
If you have any questions, don't hesitate to reach out. (zzhang42@nd.edu)