Файл:EMBEDDING ENTITIES AND RELATIONS FOR LEARNING AND INFERENCE IN KNOWLEDGE BASES 1412.6575v4.pdf
BishanYang1˚,Wen-tauYih2,XiaodongHe2,JianfengGao2 &LiDeng2 1Department of Computer Science, Cornell University, Ithaca, NY, 14850, USA bishan@cs.cornell.edu 2Microsoft Research, Redmond, WA 98052, USA {scottyih,xiaohe,jfgao,deng}@microsoft.com
ABSTRACT
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. Under this framework, we compare a variety of embedding models on the link prediction task. We show that a simple bilinear formulation achieves new state-of-the-art results for the task (achieving a top-10 accuracy of 73.2% vs. 54.7% by TransE on Freebase). Furthermore, we introduce a novel approach that utilizes the learned relation embeddings to mine logical rules such as BornInCitypa,bq^ CityInCountrypb,cq ùñ Nationalitypa,cq. We find that embeddings learned from the bilinear objective are particularly good at capturing relational semantics, and that the composition of relations is characterized by matrix multiplication. More interestingly, we demonstrate that our embedding-based rule extraction approach successfully outperforms a state-ofthe-art confidence-based rule mining approach in mining Horn rules that involve compositional reasoning.
История файла
Нажмите на дату/время, чтобы просмотреть, как тогда выглядел файл.
Дата/время | Размеры | Участник | Примечание | |
---|---|---|---|---|
текущий | 16:40, 22 декабря 2016 | 0 × 0 (1,43 МБ) | Slikos (обсуждение | вклад) |
- Вы не можете перезаписать этот файл.
Использование файла
Следующая 1 страница ссылается на данный файл: