Knowledge graph completion(KGC)can solve the problem of data sparsity in the knowledge graph.A large number of models for the KGC task have been proposed in recent years.However,the underutilisa-tion of the structure information around nodes is one of the main problems of the previous KGC model,which leads to relatively single encoding information.To this end,a new KGC model that encodes and decodes the fea-ture information is proposed.First,we adopt the sub-graph sampling method to extract node structure.Moreover,the graph convolutional network(GCN)intro-duced the channel attention convolution encode node structure features and represent them in matrix form to fully mine the node feature information.Eventually,the high-dimensional structure analysis weight decodes the encoded matrix embeddings and then constructs the scor-ing function.The experimental results show that the model performs well on the datasets used.

NIU Haoran;HE Haitao;FENG Jianzhou;NIE Junlan;ZHANG Yangsen;REN Jiadong

Yanshan University,Qinhuangdao 066004,China;Beijing Information Science and Technology University,Beijing 100096,China

电子学报（英文）

[1]NIU Haoran;HE Haitao;FENG Jianzhou;NIE Junlan;ZHANG Yangsen;REN Jiadong-.Knowledge Graph Completion Based on GCN of Multi-Information Fusion and High-Dimensional Structure Analysis Weight)[J].电子学报（英文）,2022(02):387-396

underutilisa,decodes,scor

Knowledge,Graph,Completion,Based,GCN,Multi,Information,Fusion,High,Dimensional,Structure,Analysis,Weight,graph,completion,KGC,can,solve,sparsity,knowledge,large,number,models,task,have,been,proposed,recent,years,However,structure,information,around,nodes,one,main,problems,previous,which,leads,relatively,single,encoding,To,this,end,new,that,encodes,First,adopt,sub,sampling,method,extract,Moreover,convolutional,network,intro,duced,channel,attention,features,represent,them,matrix,fully,mine,Eventually,high,dimensional,analysis,weight,encoded,embeddings,then,constructs,function,experimental,results,show,performs,well,datasets,used

0.577317