首页 > 其他分享> > [paper reading][EMNLP 2018] Learning Sequence Encoders for Temporal Knowledge Graph Completion
[paper reading][EMNLP 2018] Learning Sequence Encoders for Temporal Knowledge Graph Completion
作者:互联网
- EMNLP 2018
- https://arxiv.org/pdf/1809.03202.pdf
- link prediction in knowledge graphs
- static multi-relational data? hold for a time!
- LSTM: a sequence for a predicate (with timestamp)
1 Introduction
- triplet
- completion: s, p, ?; ?, p, o
- time info
- s, p, o, time predicate, timestamp
- sparse, irregular
3 Time-Aware Representations
- learn a scoring function \(f\)
- embedding of s,p,o
- then \(f(s,p,o)\)
- TransE, distMult
- \(p_{seq}\)!
- LSTMs
标签:Completion,predicate,Encoders,Knowledge,timestamp,EMNLP,2018,time,pdf 来源: https://www.cnblogs.com/minor-second/p/15522617.html