NSL OpenIR  > 中国科学院成都文献情报中心  > 信息技术部
一种改进的基于TransE的知识图谱表示方法
陈文杰; 文奕; 张鑫; 杨宁; 赵爽
2019-08-08
Source Publication计算机工程
Volume3Issue:23Pages:14
Abstract

自从基于距离的表示学习模型TransE被提出后,一系列模型对TransE进行改进和补充,如TransH、TransG等。然而,这类基于距离的翻译模型不能很好的处理一对多,多对一和多对多关系,而且往往孤立的学习三元组,没有考虑到知识图谱的网络结构和语义信息。本文提出了TransGraph,该模型基于TransE能够同时学习三元组和知识图谱网络结构的特征,进一步增强了知识图谱的表示效果;为了实现网络结构信息和三元组信息的深度融合,提出了一个向量共享的交叉训练机制。实验表明,相比TransE模型,TransGraph在链路预测和三元组分类等任务的各项指标上取得了显著的提升。

Other Abstract

Since the distance-based model TransE has been proposed, a series of models try to improve TransE, such as TransH and TransG. But distance-based models could not deal well with one-to-many, many-to-one, and many-to-many relationships, and often isolated learning triples do not take into account the network structure and semantic information of the knowledge graph. TransGraph based on TransE ,which simultaneously learn triples and knowledge graph network structure. In order to realize the deep fusion of network structure information and triplet information, a vector-sharing cross-training mechanism is proposed. The results show that TransGraph has achieved significant improvements in link prediction and triplet classification compared to the TransE.

Keyword知识图谱 表示学习 transe Transgraph 神经网络
Indexed ByCSCD
Language中文
Citation statistics
Document Type期刊论文
Identifierhttp://ir.las.ac.cn/handle/12502/10738
Collection中国科学院成都文献情报中心_信息技术部
Affiliation中国科学院成都文献情报中心
First Author Affilication中国科学院文献情报中心
Recommended Citation
GB/T 7714
陈文杰,文奕,张鑫,等. 一种改进的基于TransE的知识图谱表示方法[J]. 计算机工程,2019,3(23):14.
APA 陈文杰,文奕,张鑫,杨宁,&赵爽.(2019).一种改进的基于TransE的知识图谱表示方法.计算机工程,3(23),14.
MLA 陈文杰,et al."一种改进的基于TransE的知识图谱表示方法".计算机工程 3.23(2019):14.
Files in This Item: Download All
File Name/Size DocType Version Access License
一种改进的基于TransE的知识图谱表示(474KB)期刊论文作者接受稿开放获取CC BY-NC-SAView Download
Related Services
Recommend this item
Bookmark
Usage statistics
Export to Endnote
Google Scholar
Similar articles in Google Scholar
[陈文杰]'s Articles
[文奕]'s Articles
[张鑫]'s Articles
Baidu academic
Similar articles in Baidu academic
[陈文杰]'s Articles
[文奕]'s Articles
[张鑫]'s Articles
Bing Scholar
Similar articles in Bing Scholar
[陈文杰]'s Articles
[文奕]'s Articles
[张鑫]'s Articles
Terms of Use
No data!
Social Bookmark/Share
File name: 一种改进的基于TransE的知识图谱表示方法.docx
Format: Microsoft Word
All comments (0)
No comment.
 

Items in the repository are protected by copyright, with all rights reserved, unless otherwise indicated.