Modeling Heterogeneous Hierarchies with Relation-specific Hyperbolic Cones
Modeling Heterogeneous Hierarchies with Relation-specific Hyperbolic Cones
复制标题
DOI：
--
复制
发表时间：
2021-10
期刊：
影响因子：
--
通讯作者：
Yushi Bai;Rex Ying;Hongyu Ren;J. Leskovec
中科院分区：
文献类型：
--
作者：
Yushi Bai;Rex Ying;Hongyu Ren;J. Leskovec
层次关系对于组织由知识图（KG）捕获的人类知识是普遍的和不可或缺的。层次关系的关键属性是它们在实体上引入了偏序，这需要被建模以允许层次推理。然而，目前的KG嵌入只能对单个全局层次结构（单个全局偏序）进行建模，并且无法对单个KG中存在的多个异构层次结构进行建模。在这里，我们提出了ConE（Cone Embedding），这是一种KG嵌入模型，能够同时对知识图中的多个层次和非层次关系进行建模。ConE将实体嵌入到双曲锥中，并将关系建模为锥之间的转换。特别地，锥E在双曲嵌入空间的不同子空间中使用锥包含约束来捕获多个异构层次。在标准知识图基准测试上的实验表明，ConE在层次推理任务和层次图上的知识图完成任务上获得了最先进的性能。特别是，我们的方法在WN18RR上产生了45.3%的新的最先进的命中率，在DDB14上产生了16.1%的命中率（0.231 MRR）。对于层次推理任务，我们的方法在三个数据集上的平均性能比以前的最佳结果高出20%。
Hierarchical relations are prevalent and indispensable for organizing human knowledge captured by a knowledge graph (KG). The key property of hierarchical relations is that they induce a partial ordering over the entities, which needs to be modeled in order to allow for hierarchical reasoning. However, current KG embeddings can model only a single global hierarchy (single global partial ordering) and fail to model multiple heterogeneous hierarchies that exist in a single KG. Here we present ConE (Cone Embedding), a KG embedding model that is able to simultaneously model multiple hierarchical as well as non-hierarchical relations in a knowledge graph. ConE embeds entities into hyperbolic cones and models relations as transformations between the cones. In particular, ConE uses cone containment constraints in different subspaces of the hyperbolic embedding space to capture multiple heterogeneous hierarchies. Experiments on standard knowledge graph benchmarks show that ConE obtains state-of-the-art performance on hierarchical reasoning tasks as well as knowledge graph completion task on hierarchical graphs. In particular, our approach yields new state-of-the-art Hits@1 of 45.3% on WN18RR and 16.1% on DDB14 (0.231 MRR). As for hierarchical reasoning task, our approach outperforms previous best results by an average of 20% across the three datasets.