论文链接:https://pdfs.semanticscholar.org/2a3f/862199883ceff5e3c74126f0c80770653e05.pdf
论文题目:Knowledge Graph embedding by Translating on Hyperplanes
abstract:
论文动机:We discuss some mapping properties of relations which should be considered in embedding, such as reflexive, one-to-many, many-to-one, and many-to-many. We note that TransE does not do well in dealing with these properties.
文中提出应该将关系进行映射,因为考虑到transE只能表示一对一的关系,对于一对多,多对一,多对多关系并没有考虑进去,比如(h1,r1,t1)是数据集中的一个positive样本,(t1,r1,h1)也是数据集中的positive样本,那么最后transE会得到h1=t1 and r1=0;
再比如多对一或者多对一关系中:(h1,r1,t1),(h2,r1,t1)…(hn,r1,t1),then transE会得到h1=h2=…=hn
introduction:
Accordingly we propose a method named translation on hyperplanes (TransH) which interprets a relation as a translating operation on a hyperp
关于知识图谱的trans系列二:transH
最新推荐文章于 2023-09-02 03:58:13 发布
TransH模型为解决TransE在处理一对多、多对一和多对多关系上的不足,提出了将关系视为超平面上的转换操作。每个关系由超平面规范向量wr和超平面上的转换向量dr描述。实体在超平面上的映射解决了过度收敛问题,优化了负样本构造,以提高模型预测准确性。评价指标包括meanRank和hit@10。
摘要由CSDN通过智能技术生成