Knowledge Graph Embedding via Dynamic Mapping Matrix
By Guoliang Ji, Shizhu He, Liheng Xu, Kang Liu, Jun Zhao. 2015
1,714 citations recorded by the source corpus.
Knowledge graphs are useful resources for numerous AI applications, but they are far from completeness. Previous work such as TransE, TransH and TransR/CTransR regard a relation as translation from head entity to tail entity and the CTransR achieves state-of-the-art performance. In this paper, we propose a more fine-grained model named TransD, which is an improvement of TransR/CTransR. In TransD, we use two vectors to represent a named symbol object (entity and relation). The first one represents the meaning of a(n) entity (relation), the other one is used to construct mapping matrix dynamically. Compared with TransR/CTransR, TransD not only considers the diversity of relations, but also entities. TransD has less parameters and has no matrix-vector multiplication operations, which makes it can be applied on large scale graphs. In Experiments, we evaluate our model on two typical tasks including triplets classification and link prediction. Evaluation results show that our approach outperforms state-of-the-art methods.
Subfields: Artificial Intelligence, Computer Vision and Pattern Recognition, Graph Learning, Knowledge Graphs & Representation
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
- Zhiyuan Liu
- Jianfeng Gao
- Mohit Bansal
- Maosong Sun
- Minlie Huang
- Erik Cambria
- Percy Liang
- Noah A. Smith
In Computer Vision and Pattern Recognition
- Luc Van Gool
- Dacheng Tao
- Yi Yang
- Jianfeng Gao
- Trevor Darrell
- Xiaogang Wang
- Ming-Hsuan Yang
- Andrea Vedaldi
In Graph Learning
- Zhiyuan Liu
- Maosong Sun
- William Yang Wang
- Minlie Huang
- Xiaoyan Zhu
- Xing Xie
- Avishek Joey Bose
- William L. Hamilton
In Knowledge Graphs & Representation
Citations in the corpus
- RippleNet: Propagating User Preferences on the Knowledge Graph for Recommender Systems (2018, methodology): “addition, our work also incorporates a KGE term as a regularizer for more stable and effective learning. 3.6.3 Knowledge Graph Embedding. RippleNet also connects to a large body of work in KGE methods[3, 10, 14, 19, 28, 30, 37, 41]. KGE intends to embed entities and relations in a KG into continuous”
- DKN: Deep Knowledge-Aware Network for News Recommendation (2018, methodology): “• TransD [18] replaces the projection matrix in TransR by the product of two projection vectors of an entity-relation pair: where h ⊥ = ( r p h ⊤ p + I ) h , t ⊥ = ( r p t ⊤ p + I ) t , h p , r p and t p are another set of vectors for entities and relations, and I is the identity matrix.”
- Compositional Fairness Constraints for Graph Embeddings (2019, methodology): “In all experiments, we used multi-layer perceptrons (MLPs) with leaky ReLU activation functions (Xu et al., 2015) as the discriminators D k and filters f k .”
- A Comprehensive Survey of Graph Embedding: Problems, Techniques, and Applications (2017, neutral): “TransD [125] kMrhhþ r Mrttk22 TranSparse [126] kM r ður Þhþ r M rðurÞtk 2 l1=2 m-TransH [127] k P r2MðRrÞ arðrÞPnr ðtðrÞÞ þ brk (2); t 2 NMðRrÞ”
- Recent advances in deep learning based dialogue systems: a systematic survey (2021, positive): “Models such as TransE (Bordes et al., 2013), TransR (Lin et al., 2015), TransH (Wang et al., 2014), TransD (Ji et al., 2015), TransG (Xiao et al., 2015), TransM (Fan et al., 2014), HolE (Nickel et al., 2016) and ProjE (Shi and Weninger, 2017) belong to this category.”
- A Comprehensive Survey of Scene Graphs: Generation and Application (2021, methodology): “Unlike TransE- based SGG methods, RLSV uses TransD [51] (an improved version of TransR/CTransR [47]) to project the entities (subjects and objects) from the entity space to the relation space by means of two mapping matrices (Ms and Mo), and the triplets hs p oi have new representations hs ? p ? o ?”
- Tensor Decompositions for temporal knowledge base completion (2020, methodology): “This method has lead to many variations (Ji et al., 2015; Nguyen et al., 2016; Wang et al., 2014), but is limited in the relation systems it can model (Kazemi & Poole, 2018) and does not lead to state of the art performances on current benchmarks.”
- DeepPath: A Reinforcement Learning Method for Knowledge Graph Reasoning (2017, methodology): “For embedding based methods, we evaluate several state-of-the-art embeddings designed for knowledge base completion, such as TransE (Bordes et al., 2013), TransH (Wang et al., 2014), TransR (Lin et al., 2015) and TransD (Ji et al., 2015) .”
- 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.”
- A Survey on Knowledge Graphs: Representation, Acquisition, and Applications (2020, neutral): “TransD [28] h, t,whwt ∈ R r,wr ∈ R − ∥∥∥(wrw> h + I)h + r− (wrw> t + I) t∥∥∥2 2 TransM [173] h, t ∈ R r ∈ R −θr‖h + r− t‖1/2”
- Knowledge-Driven Event Embedding for Stock Prediction (2016, methodology): “Knowledge Graph Embedding Recently, several methods have been explored to represent and encode knowledge graph (Bordes et al., 2013; Bordes et al., 2014; Chang et al., 2013; Ji et al., 2015; Lin et al., 2015) in distributed vectors.”
- Prototypical Representation Learning for Relation Extraction (2021, background): “Although many works are proposed to learn representations for relations from well-structured knowledge (Bordes et al., 2013; Lin et al., 2015; Ji et al., 2015), when we extend the learning source to be unstructured distantly-labeled text (Mintz et al., 2009), this task becomes particularly…”
- Analogical Inference for Multi-relational Embeddings (2017, neutral): “Unstructured (Bordes et al., 2013) 38.2 6.3 RESCAL (Nickel et al., 2011) 52.8 44.1 NTN (Socher et al., 2013) 66.1 41.4 SME (Bordes et al., 2012) 74.1 41.3 SE (Bordes et al., 2011) 80.5 39.8 LFM (Jenatton et al., 2012) 81.6 33.1 TransH (Wang et al., 2014) 86.7 64.4 TransE (Bordes et al., 2013) 89.2 4”
- Distant Supervision for Relation Extraction with Sentence-Level Attention and Entity Descriptions (2017, background): “Attention Layer Recently, many knowledge graph embedding approaches (Bordes et al. 2013; Wang et al. 2014; Ji et al. 2015) regarded relation as translation from head entity (e1) to tail entity (e2).”
- Large-scale Multi-modal Pre-trained Models: A Comprehensive Survey (2023, background): “TransR is proposed to embed entity and relation in a separated spaces to capture different aspects of entities over various relations [199].”
- Incorporating Knowledge Graph Embeddings into Topic Modeling (2017, positive): “Since cosine distance is typically used to measure similarity between entity embeddings (Ji et al. 2015; He et al. 2015; Yang et al. 2015) and some knowledge graph embeddings lie on a unit sphere ( 2 norm equals to 1) (Bordes et al. 2011; 2013; 2014; Yang et al. 2015; Garcia-Duran et al. 2016), we…”
- Deep learning for noise-tolerant RDFS reasoning (2019, methodology): “N relations such as TransH [28] which uses relation-specific hyperplanes, TransR [55] and TransD [56].”
- Cross-lingual Knowledge Graph Alignment via Graph Convolutional Networks (2018, methodology): “To further improve TransE, several enhanced models based on it have been proposed, including TransR (Lin et al., 2015), TransH (Wang et al., 2014) and TransD (Ji et al., 2015) etc.”
- Representation Learning of Knowledge Graphs with Hierarchical Types (2016, methodology): “[Krompaß et al., 2015] considers entity types as hard constraints in latent variable models for KGs.”