lrlbbzl/presa

Prior Relational Schema Assists Effective Contrastive Learning for Inductive Knowledge Graph Completion

Code of COLING 2024 paper Prior Relational Schema Assists Effective Contrastive Learning for Inductive Knowledge Graph Completion.

In this article, we propose an efficient KGE method based on relational schema and contrastive learning for inductive knowledge graph completion tasks.

Requirements

• python>=3.7
• torch>=1.6
• transformers>=4.15

We have run the experiments on 4 * A40 (when using high batch size).

Tutorial

Humanswiki-ind

• Training

  bash scripts/train_hw.sh

  Modify parameters such as CUDA_VISIBLE_DEVICES according to your own needs.

• Inference

  bash scripts/eval.sh ./checkpoint/humans_wikidata_ind/model_last.mdl humans_wikidata_ind

Acknowledge

Thanks the sharing from SimKGC: Simple Contrastive Knowledge Graph Completion with Pre-trained Language Models - ACL Anthology

presa's Issues

关于数据集的问题

不好意思，打扰你了，我在代码中看到你没有提供Fb15k-237和WN18RR的数据集，请问你是否方便提供相关数据集，如果提供将万分感谢，期待你的回复