Embedding Entities and Relations for Learning and Inference in Knowledge Bases

20 Dec 2014 Bishan Yang Wen-tau Yih Xiaodong He Jianfeng Gao Li Deng

We consider learning representations of entities and relations in KBs using the neural-embedding approach. We show that most existing models, including NTN (Socher et al., 2013) and TransE (Bordes et al., 2013b), can be generalized under a unified learning framework, where entities are low-dimensional vectors learned from a neural network and relations are bilinear and/or linear mapping functions... (read more)

PDF Abstract

Results from the Paper


TASK DATASET MODEL METRIC NAME METRIC VALUE GLOBAL RANK RESULT BENCHMARK
Link Prediction FB15k-237 DisMult MRR 0.241 # 39
Hits@10 0.419 # 41
Link Prediction WN18 DistMult MRR 0.822 # 21
Hits@10 0.936 # 24
Hits@3 0.914 # 16
Hits@1 0.728 # 17
MR 902 # 16
Link Prediction WN18RR DisMult MRR 0.43 # 34
Hits@1 0.39 # 28

Methods used in the Paper


METHOD TYPE
TransE
Graph Embeddings