Link prediction based on higher-order structure extraction and autoencoder learning in directed networks

作者：

Highlights：

• We propose a node embedding method that considers the higher-order structure.

• We propose an information aggregation method based on attention mechanism.

• We propose an efficient decoder scheme to predict the directions of links.

• We provide an autoencoder based link prediction method for directed networks.

摘要

•We propose a node embedding method that considers the higher-order structure.•We propose an information aggregation method based on attention mechanism.•We propose an efficient decoder scheme to predict the directions of links.•We provide an autoencoder based link prediction method for directed networks.

论文关键词：Link prediction,Directed networks,Autoencoder,Higher-order structure,Attention,PageRank

论文评审过程：Received 28 June 2021, Revised 4 January 2022, Accepted 18 January 2022, Available online 25 January 2022, Version of Record 12 February 2022.

论文官网地址：https://doi.org/10.1016/j.knosys.2022.108241