Efficient Hyper-parameter Search for Knowledge Graph Embedding

Yongqi Zhang, Zhanke Zhou, Quanming Yao, Yong Li


Abstract
While hyper-parameters (HPs) are important for knowledge graph (KG) learning, existing methods fail to search them efficiently. To solve this problem, we first analyze the properties of different HPs and measure the transfer ability from small subgraph to the full graph. Based on the analysis, we propose an efficient two-stage search algorithm KGTuner, which efficiently explores HP configurations on small subgraph at the first stage and transfers the top-performed configurations for fine-tuning on the large full graph at the second stage. Experiments show that our method can consistently find better HPs than the baseline algorithms within the same time budget, which achieves 9.1% average relative improvement for four embedding models on the large-scale KGs in open graph benchmark. Our code is released in https://github. com/AutoML-Research/KGTuner.
Anthology ID:
2022.acl-long.194
Volume:
Proceedings of the 60th Annual Meeting of the Association for Computational Linguistics (Volume 1: Long Papers)
Month:
May
Year:
2022
Address:
Dublin, Ireland
Venue:
ACL
SIG:
Publisher:
Association for Computational Linguistics
Note:
Pages:
2715–2735
Language:
URL:
https://aclanthology.org/2022.acl-long.194
DOI:
10.18653/v1/2022.acl-long.194
Bibkey:
Cite (ACL):
Yongqi Zhang, Zhanke Zhou, Quanming Yao, and Yong Li. 2022. Efficient Hyper-parameter Search for Knowledge Graph Embedding. In Proceedings of the 60th Annual Meeting of the Association for Computational Linguistics (Volume 1: Long Papers), pages 2715–2735, Dublin, Ireland. Association for Computational Linguistics.
Cite (Informal):
Efficient Hyper-parameter Search for Knowledge Graph Embedding (Zhang et al., ACL 2022)
Copy Citation:
PDF:
https://aclanthology.org/2022.acl-long.194.pdf
Code
 automl-research/kgtuner
Data
FB15k-237OGB