Knowledge Graph Embedding by Translating on Hyperplanes
By Zhen Wang, Jianwen Zhang, Jianlin Feng, Zheng Chen. 2014
4,058 citations recorded by the source corpus.
We deal with embedding a large scale knowledge graph composed of entities and relations into a continuous vector space. TransE is a promising method proposed recently, which is very efficient while achieving state-of-the-art predictive performance. 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. Some complex models are capable of preserving these mapping properties but sacrifice efficiency in the process. To make a good trade-off between model capacity and efficiency, in this paper we propose TransH which models a relation as a hyperplane together with a translation operation on it. In this way, we can well preserve the above mapping properties of relations with almost the same model complexity of TransE. Additionally, as a practical knowledge graph is often far from completed, how to construct negative examples to reduce false negative labels in training is very important. Utilizing the one-to-many/many-to-one mapping property of a relation, we propose a simple trick to reduce the possibility of false negative labeling. We conduct extensive experiments on link prediction, triplet classification and fact extraction on benchmark datasets like WordNet and Freebase. Experiments show TransH delivers significant improvements over TransE on predictive accuracy with comparable capability to scale up.
Subfields: Artificial Intelligence, Statistical and Nonlinear Physics, Graph Learning
How researchers see this paper
The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.
In Artificial Intelligence
- Mohit Bansal
- Jianfeng Gao
- Zhiyuan Liu
- Luke S. Zettlemoyer
- Percy Liang
- Maosong Sun
- Noah A. Smith
- Yejin Choi
In Statistical and Nonlinear Physics
In Graph Learning
- Zhiyuan Liu
- Maosong Sun
- Tim Weninger
- Baoxu Shi
- William L. Hamilton
- Minlie Huang
- Xiaoyan Zhu
- Xing Xie
Citations in the corpus
- Knowledge Enhanced Contextual Word Representations (2019, methodology): “These methods broadly fall into two categories: translational distance models (Bordes et al., 2013; Wang et al., 2014b; Lin et al., 2015; Xiao et al., 2016) which use a distance-based scoring function, and linear models (Nickel et al., 2011; Yang et al., 2014; Trouillon et al., 2016; Dettmers et…”
- RippleNet: Propagating User Preferences on the Knowledge Graph for Recommender Systems (2018, methodology): “its inherent structure. Readers can refer to [30] for a more comprehensive survey. KGE methods are mainly classified into two categories: (1) translational distance models, such as TransE [3], TransH [32], and TransR [13], exploit distance-based scoring functions when learning representations of ent”
- DKN: Deep Knowledge-Aware Network for News Recommendation (2018, methodology): “In addition to translation-based methods [4, 18, 26, 48] used in DKN, researchers have also proposed many other models such as Structured Embedding [5], Latent Factor Model [17], Neural Tensor Network [37] and GraphGAN [42].”
- Learning Entity and Relation Embeddings for Knowledge Graph Completion (2015, methodology): “Since the freebase part of our data set is built by ourselves, different from the ones in (Wang et al. 2014), the evaluation results cannot be compared directly with those reported in (Wang et al. 2014).”
- Knowledge Infused Learning (K-IL): Towards Deep Incorporation of Knowledge in Deep Learning (2019, methodology): “2013), TRANS-H (Wang et al. 2014), and HOLE (Nickel et al.”
- A Comprehensive Survey of Graph Embedding: Problems, Techniques, and Applications (2017, background): “TransH [14] kðh wr hwrÞ þ dr ðt wr twrÞk 2 2”
- Representing Text for Joint Embedding of Text and Knowledge Bases (2015, methodology): “The FB15k dataset has been used in multiple studies on knowledge base completion (Wang et al., 2014b; Yang et al., 2015).”
- GRAM: Graph-based Attention Model for Healthcare Representation Learning (2016, background): “They aim to project entities and relations [5, 23, 35, 40] to the latent space based on the triples or additional information such as hierarchy of entities [42].”
- Knowledge Graph Embedding via Dynamic Mapping Matrix (2015, methodology): “Therefore, we obtain ∆ from knowledge graphs and generate ∆ as follows: We also use two strategies “unif” and “bern” described in (Wang et al. 2014) to replace the head or tail entity.”
- Low-Dimensional Hyperbolic Knowledge Graph Embeddings (2020, methodology): “These include translation approaches (Bordes et al., 2013; Ji et al., 2015; Wang et al., 2014; Lin et al., 2015) or tensor factorization methods such as RESCAL (Nickel et al., 2011) or DistMult (Yang et al., 2015).”
- Relation-Aware Entity Alignment for Heterogeneous Knowledge Graphs (2019, methodology): “, 2013], TransH [Wang et al., 2014], and PTransE [Lin et al.”
- Prototypical Representation Learning for Relation Extraction (2021, neutral): “…features of KBs. -8.657383 35.28101 -13.320659 29.52109 -11.843529 26.295746 -10.6841545 28.803246 (Nickel et al., 2011; Bordes et al., 2013; et al., 2015 Wang et al., 2014; Nickel et al., 2016; Ji et al., 2015; Lin et al., 2015; Trouillon et al., 2016; Sun et al., 2019; Balaˇzevi´c et al., 2019).”
- OpenKE: An Open Toolkit for Knowledge Embedding (2018, methodology): “Scoring Function Parameters Loss Function RESCAL (Nickel et al., 2011) TransH (Wang et al., 2014) TransR (Lin et al., 2015) TransD (Ji et al., 2015) HolE (Nickel et al., 2016) r ComplEx (Trouillon et al., 2016) Table 1 : The brief introduction of some typical KE models.”
- On the Representation and Embedding of Knowledge Bases beyond Binary Relations (2016, methodology): “Recently knowledge base embedding[Bor-des et al. , 2013; Wang et al. , 2014; Lin et al. , 2015b; Bordes et al. , 2014b; 2011; Socher et al. , 2013; Lin et al. , 2015a] has stood out as an appealing and generic methodology to access various research problems in this area.”
- Composing Relationships with Translations (2015, background): “Many extensions have been proposed to improve the representation power of T RANS E while still keeping its simplicity, by adding some projections steps before the translation (Wang et al., 2014; Lin et al., 2015).”
- Large-scale Multi-modal Pre-trained Models: A Comprehensive Survey (2023, methodology): “To deal with mapping properties of relations efficiently in complex models, such as reflexive, one-to-many, many-to-one and many-to-many, TransH is proposed to model a relation as a translation operation on a hyperplane [198].”
- Semantically Smooth Knowledge Graph Embedding (2015, methodology): “…represented as points (i.e. vectors) in a continuous vector space, and relations as operators in that space, which can be characterized by vectors (Bordes et al., 2013; Bordes et al., 2014; Wang et al., 2014b), matrices (Bordes et al., 2011; Jenatton et al., 2012), or tensors (Socher et al., 2013).”