On Approximate Reasoning Capabilities of Low-Rank Vector Spaces

By Guillaume Bouchard, Sameer Singh, Théo Trouillon. 2015

98 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.

Citations in the corpus

Integrating Learning and Reasoning with Deep Logic Models (2019, background): “The Countries dataset [5] consists of 244 countries (e.”
Complex Embeddings for Simple Link Prediction (2016, methodology): “Dot products of embeddings scale well and can naturally handle both symmetry and (ir-)reﬂexivity of relations; us-ing an appropriate loss function even enables transitivity (Bouchard et al., 2015).”
End-to-end Differentiable Proving (2017, background): “Countries The Countries KB is a dataset introduced by [32] for testing reasoning capabilities of neural link prediction models.”
Holographic Embeddings of Knowledge Graphs (2015, methodology): “This can, for instance, be done by minimizing the (regularized) logistic loss (2) For relational data, minimizing the logistic loss has the additional advantage that it can help to ﬁnd low dimensional embeddings for complex relational patterns (Bouchard, Singh, and Trouillon, 2015).”
Convolutional 2D Knowledge Graph Embeddings (2017, background): “It was first noted by Toutanova and Chen (2015) that WN18 and FB15k suffer from test leakage through inverse relations: a large number of test triples can be obtained simply by inverting triples in the training set. For example, the test set frequently contains triples such as (s, hyponym, o) while ”
Poincaré Embeddings for Learning Hierarchical Representations (2017, background): “Although non-linear embeddings can mitigate this problem [8], complex graph patterns can still require a computationally infeasible embedding dimension.”
Knowledge Graph Embedding: A Survey of Approaches and Applications (2017, background): “It has been shown that minimizing the logistic loss can help to find compact representations for some complex relational patterns such as transitive relations [73].”
RotatE: Knowledge Graph Embedding by Relational Rotation in Complex Space (2018, methodology): “00 Table 6: Results on the Countries datasets.”
Inference is Everything: Recasting Semantic Resources into a Unified Evaluation Framework (2017, background): “Vector space-based intermediate forms are not commonly recognized as logical forms but in light of recent work (Bouchard et al., 2015) it seems worthwhile to reconsider this view.”
Knowledge Graph Completion via Complex Tensor Factorization (2017, methodology): “, 2014), or only in specific cases (Bouchard et al., 2015).”
A Survey of Knowledge Graph Reasoning on Graph Types: Static, Dynamic, and Multi-Modal (2022, background): “Facts # Test Facts ATOMIC [205] 304,388 9 610,536 87,700 87,701 Countries [206] 271 2 1,110 24 24 CoDEX-S [207] 2,034 42 32,888 3,654 3656 CoDEX-M [207] 17,050 51 185,584 20620 20622 CoDEX-L [207] 77,951 69 551,193 30,622 30622 ConceptNet [208] 28,370,083 50 27,259,933 3,407,492 3,407,492 ConceptNet”
Differentiable Reasoning on Large Knowledge Bases and Natural Language (2019, unknown): “We can also see that models learned by GNTPs are interpretable : in Table 1 Figure 3: GNTPs on Countries with generated mentions.”
Go for a Walk and Arrive at the Answer: Reasoning Over Paths in Knowledge Bases using Reinforcement Learning (2017, methodology): “Dataset We use three standard datasets: COUNTRIES (Bouchard et al., 2015), KINSHIP, and UMLS (Kok & Domingos, 2007).”