面向知识图谱归纳链接预测的负采样方法  

A Negative Sampling Method for Knowledge Graph Inductive Link Prediction

作　　者：刘洪波[1] 陈越[1] 杨奎武[1] 吴皓 张潮 LIU Hongbo;CHEN Yue;YANG Kuiwu;WU Hao;ZHANG Chao(Information Engineering University,Zhengzhou 450001,China)

机构地区：[1]信息工程大学,河南郑州450001

出　　处：《信息工程大学学报》2025年第2期142-147,160,共7页Journal of Information Engineering University

基　　金：国家自然科学基金(62172433)。

摘　　要：知识图谱归纳链接预测模型在训练过程中需要使用正例三元组和负例三元组,而当前的随机负采样方法容易产生低质量的负例三元组,影响模型的特征学习能力。针对该问题,提出一种基于相似性的负采样方法。首先,获取被替换实体周围的k跳邻居节点集合;其次,从该集合中挑选相似度高的实体替换原三元组中头实体或者尾实体,从而生成负例三元组;最后,将该方法应用在归纳链接预测模型中,并在WN18RR和FB15K-237数据集上进行归纳链接预测实验。实验结果表明,相比其他模型,该方法在MRR指标最高提升10.47个百分点,在Hits@10指标最高提升16.02个百分点。通过负样本质量分析,进一步说明该负采样方法生成高质量的负例三元组,提升模型的性能。The inductive link prediction model of knowledge graphs requires both positive and negative triplets during the training process.However,the current random negative sampling method tends to produce low-quality negative triplets,which affects the feature learning ability of the model.To address this problem,a similarity-based negative sampling method is proposed.Firstly,the set of k-hop neighbor nodes around the replaced entity is obtained.Secondly,the entities with high similarity are selected from the set to replace the head or tail entities in the original triplet,so as to generate a negative triplet.Finally,the negative sampling method is used in the inductive link prediction model,and the inductive link prediction experiments are carried out on the WN18RR and FB15K-237 datasets.Experimental results demonstrate that compared with other models,the MRR metric is increased by up to 10.47 percentage point,and the Hits@10 metric is increased by up to 16.02 percentage point.Furthermore,the negative sample quality analysis illustrates that high-quality negative triplets are generated by using the negative sampling method,which improves the performance of the model.

关 键 词：知识图谱 归纳链接预测 负采样 k跳邻居 相似性计算 

分 类 号：TP391[自动化与计算机技术—计算机应用技术]