Learning flexible network representation via anonymous walks

作者：

Highlights：

• Anonymous walks can capture local structural patterns.

• Incorporating local structural information into the learning process.

• Proposing two strategies: statistic-based model and embedding-based model.

• “Anonymous Walks Vectors” can be viewed as “Paragraph Vectors” on the graph.

摘要

•Anonymous walks can capture local structural patterns.•Incorporating local structural information into the learning process.•Proposing two strategies: statistic-based model and embedding-based model.•“Anonymous Walks Vectors” can be viewed as “Paragraph Vectors” on the graph.

论文关键词：Network analysis,Network representation learning,Local structural patterns,Anonymous walks

论文评审过程：Received 29 September 2020, Revised 1 April 2021, Accepted 3 April 2021, Available online 5 April 2021, Version of Record 15 April 2021.

论文官网地址：https://doi.org/10.1016/j.knosys.2021.107021