GrAR: A novel framework for Graph Alignment based on Relativity concept

作者：

Highlights：

• GrAR is introduced to add role information to the feature vectors.

• Relative-Degree of node is the degree of node relative to the degree of its neighbors.

• Neighbors with higher relative degrees are more resistant to noise.

• REGAL is not a framework and cannot increase accuracy of ID vector.

摘要

•GrAR is introduced to add role information to the feature vectors.•Relative-Degree of node is the degree of node relative to the degree of its neighbors.•Neighbors with higher relative degrees are more resistant to noise.•REGAL is not a framework and cannot increase accuracy of ID vector.

论文关键词：GrAR,Relativity concept,Graph alignment,Social Network Analysis,Anchor Link,Entity Resolution

论文评审过程：Received 19 April 2020, Revised 15 April 2021, Accepted 11 September 2021, Available online 17 September 2021, Version of Record 22 September 2021.

论文官网地址：https://doi.org/10.1016/j.eswa.2021.115908