Dynamic Heterogeneous Graph Attention Neural Architecture Search

Authors: Zeyang Zhang, Ziwei Zhang, Xin Wang, Yijian Qin, Zhou Qin, Wenwu Zhu

AAAI 2023 | Conference PDF | Archive PDF | Plain Text | LLM Run Details

Reproducibility Variable Result LLM Response
Research Type Experimental Extensive experiments on real-world dynamic heterogeneous graph datasets demonstrate that our proposed method significantly outperforms state-of-the-art baselines for tasks including link prediction, node classification and node regression.
Researcher Affiliation Collaboration Zeyang Zhang1*, Ziwei Zhang1, Xin Wang1 , Yijian Qin1, Zhou Qin2, Wenwu Zhu1 1Tsinghua University 2Alibaba Group
Pseudocode No The paper describes methods and algorithms textually but does not include a formally labeled pseudocode or algorithm block.
Open Source Code Yes The codes are publicly available1. 1https://github.com/wondergo2017/DHGAS
Open Datasets Yes We conduct experiments for the link prediction task on two datasets: an academic citation dataset Aminer (Ji et al. 2021) and a recommendation dataset Ecomm (Xue et al. 2020)... adopting two datasets: a business review dataset Yelp (Ji et al. 2021) and an e-commerce risk management dataset Drugs3. For the node regression task, we adopt an epidemic disease dataset COVID-19 (Fan et al. 2022).
Dataset Splits No The paper does not provide specific train/validation/test dataset split percentages or absolute counts for each split.
Hardware Specification No No specific hardware details (e.g., GPU/CPU models, processor types, memory amounts, or detailed computer specifications) used for running experiments are provided.
Software Dependencies No The paper does not provide specific software dependencies with version numbers (e.g., library or solver names with version numbers).
Experiment Setup Yes We report the results on the Aminer dataset when the localization constraint hyperparameters KLo is chosen from {4, 8, 10, 20, 40}