摘要
根据图上节点所在位置与邻居节点特征,可以使用不同策略为每个图上节点进行区间编码,基于区间编码,许多在大型图上的应用如知识图谱查询、智能问答等的处理可以加速或得到准确性上的提升。针对此种情况,提出一种基于树分解算法的图上点区间编码方法,并在大型知识图谱上通过智能问答歧义消除的应用验证该方法的有效性。实验结果表明,该方法能够有效地表达出图上节点的位置特征,并帮助智能问答中的实体消除歧义。
Based on the position and neighbors of one vertex in a graph,there exist different strategies which are able to conduct range encoding for each vertex in the graph.With range codes,plenty of applications over large graphs such as query processing and question answering over knowledge graphs can be accelerated or processed more accurately.In view of this,we propose a range encoding approach based on tree decomposition algorithm.We verified the effectiveness of the proposed approach by conducting experiments of disambiguation during question answering over large knowledge graph.Experimental results show that the proposed approach can present the position features of vertexes effectively,and help disambiguation during question answering.
作者
陈子轩
何震瀛
荆一楠
Chen Zixuan;He Zhenying;Jing Yi’nan(School of Software,Fudan University,Shanghai 201203,China;School of Computer Science,Fudan University,Shanghai 201203,China)
出处
《计算机应用与软件》
北大核心
2022年第3期254-259,共6页
Computer Applications and Software
基金
国家重点研发计划项目(2018YFB1004404)。
关键词
知识图谱
树分解
区间编码
Knowlede graph
Tree decomposition
Range encoding
