neair-neighbor_aware_inductive_reasoning_for_kgc

Published: 5 December 2025| Version 1 | DOI: 10.17632/vjxhh2n9jy.1
Contributor:
Shenhao Xiong

Description

paper code

Files

Steps to reproduce

1. install following python packages for python>=3.9: torch >= 2.0.1 scipy 2. Quick preprocess datasets for getting numpy triplets and evaluation info: python preprocess.py -d data_ind/fb237_v1 -i python preprocess.py -d data_ind/WN18RR_v1 -i python preprocess.py -d data_ind/nell_v1 -i python preprocess.py -d data/WN18RR python preprocess.py -d data/nell # run for all datasets bash scripts/preprocess_all.sh 3. Simply run following commands in scripts to reproduce: # for inductive bash scripts/FB15k237.sh v1 --device 'cuda' # for transductive bash scripts/tran.sh WN18RR --device 'cuda' view bash scripts or the entrypoint file `neair_pna_entry.py` for running config details.

Categories

Graph Neural Network, Knowledge Graph

Licence