A latent factor model for highly multi-relational data

By Rodolphe Jenatton, Nicolas Le Roux, Antoine Bordes, Guillaume Obozinski. 2012

434 citations recorded by the source corpus.

How researchers see this paper

The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.

In Graph Learning

Citations in the corpus

Recursive Deep Models for Semantic Compositionality Over a Sentiment Treebank (2013, methodology): “…to relate inputs through three way interactions, parameterized by a tensor have been proposed for relation classiﬁ-cation (Sutskever et al., 2009; Jenatton et al., 2012), extending Restricted Boltzmann machines (Ranzato and Hinton, 2010) and as a special layer for speech recognition (Yu et al.,…”
Knowledge Graph Embedding by Translating on Hyperplanes (2014, background): “Distant (Bordes et al. 2011) W rh h − Wrtt1, W rh , Wrt ∈ R k×k O(nek + 2nrk 2 ) Bilinear (Jenatton et al. 2012) h Wrt, Wr ∈ R k×k O(nek + nrk 2 ) Single Layer u r f (W rh h + Wrtt + br) O(nek + nr(sk + s)) ur, br ∈ R s , W rh , Wrt ∈ R s×k NTN (Socher et al. 2013) u r f (h Wrt + W rh h + Wrtt + br)”
DKN: Deep Knowledge-Aware Network for News Recommendation (2018, methodology): “e original graph structure. In addition to translationbased 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]. Recently, the knowledge graph has also been used”
Learning Entity and Relation Embeddings for Knowledge Graph Completion (2015, methodology): “LFM model (Jenatton et al. 2012; Sutskever, Tenenbaum, and Salakhutdinov 2009) considers second-order correlations between entity embeddings using a quadratic form, and defines a bilinear score function f r (h, t) = h M r t. Neural Tensor Network (NTN).”
Irreflexive and Hierarchical Relations as Translations (2013, methodology): “However, they have been outperformed by energy-based models (Bordes et al., 2011; Jenatton et al., 2012; Bordes et al., 2013; Chen et al., 2013).”
A semantic matching energy function for learning with multi-relational data (2013, methodology): “d evaluation metric is area under the precision-recall curve (AUC). Table 2 presents results of SME along with those of RESCAL, MRC, IRM, CP (CANDECOMP-PARAFAC) andLFM, which have been extracted from [5, 3]. The linear formulation of SME is outperformed by SME(bilinear) on all three tasks. The large”
Complex Embeddings for Simple Link Prediction (2016, methodology): “Direct extensions were based on uni-,bi-and trigram latent factors for triple data, as well as a low-rank relation matrix (Jenatton et al., 2012).”
Weakly-Supervised Learning of Visual Relations (2017, background): “Triplet learning has been addressed in various tasks such as mining typical relations (knowledge extraction) [6, 42, 51, 53], reasoning [20, 34, 44], object detection [16, 43], image retrieval [22] or fact retrieval [10].”
Embedding Entities and Relations for Learning and Inference in Knowledge Bases (2014, methodology): “Such bilinear formulation has been used in other matrix factorization models such as in (Nickel et al., 2011; Jenatton et al., 2012; Garcı́a-Durán et al., 2014) with different forms of regularization.”
Commonsense Knowledge Base Completion (2016, methodology): “We ﬁrst consider bilinear models, since they have been found useful for KBC in past work (Nickel et al., 2011; Jenatton et al., 2012; Garc´ıa-Dur´an et al., 2014; Yang et al., 2014).”
A Review of Relational Machine Learning for Knowledge Graphs (2015, background): “[81] proposed a tensor factorization model for knowledge graphs with a very large number of different relations.”
NaturalLI: Natural Logic Inference for Common Sense Reasoning (2014, methodology): “Chen et al. (2013) and Socher et al. (2013) use Neural Tensor Networks to predict unseen relation triples in WordNet and Freebase, following a line of work by Bordes et al. (2011) and Jenatton et al. (2012).”
StarSpace: Embed All The Things! (2017, methodology): “TransE was shown to out-perform RESCAL (Nickel, Tresp, and Kriegel 2011), RFM (Jenatton et al. 2012), SE (Bordes et al. 2011) and SME (Bordes et al. 2014) and is considered a standard benchmark method.”
Knowledge Graph Embedding via Dynamic Mapping Matrix (2015, background): “Similar to TransE, TransH has the score function as follows: (2) Model #Parameters # Operations (Time complexity) Unstructured (Bordes et al. 2012; 2014) O ( N e m ) O ( N t ) SE (Bordes et al. 2011) O ( N e m + 2 N r n 2 )( m = n ) O (2 m 2 N t ) SME(linear) (Bordes et al. 2012; 2014) O ( N e m + N”
Entity Embeddings of Categorical Variables (2016, background): “More recently, knowledge base which is a large collection of complex relational data is seeing lots of works using entity embedding[22–24].”
Large-Scale Embedding Learning in Heterogeneous Event Data (2016, background): “Studies of similar flavor of higher-order data [24] have recently emerged for some tasks, such as recommender system [20], multi-relational learning [25], prediction [26], and clustering [27].”
Knowledge vault: a web-scale approach to probabilistic knowledge fusion (2014, methodology): “, [31, 11, 20, 37]); and (3) methods that approximate the correlation using algorithmic approaches, such as random walks [24].”
Heterogeneous Network Embedding via Deep Architectures (2015, background): “A branch of latent feature embeddings is motivated by applications such as collaborative filtering and link prediction in networks that model the relations between entities from latent attributes [16].”
OpenKE: An Open Toolkit for Knowledge Embedding (2018, background): “…KGs into a continuous low-dimensional space, such as linear models (Bordes et al., 2011, 2012, 2014), latent factor models (Sutskever et al., 2009; Jenatton et al., 2012; Yang et al., 2015; Liu et al., 2017), neural models (Socher et al., 2013; Dong et al., 2014), matrix factorization models…”