An Unsupervised Rapid Network Alignment Framework via Network Coarsening
Network alignment aims to identify the correspondence of nodes between two or more networks. It is the cornerstone of many network mining tasks, such as cross-platform recommendation and cross-network data aggregation. Recently, with the development of network representation learning techniques, res...
| Main Authors: | , , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
MDPI AG
2023-01-01
|
| Series: | Mathematics |
| Subjects: | |
| Online Access: | https://www.mdpi.com/2227-7390/11/3/573 |
| _version_ | 1827759906145435648 |
|---|---|
| author | Lei Zhang Feng Qian Jie Chen Shu Zhao |
| author_facet | Lei Zhang Feng Qian Jie Chen Shu Zhao |
| author_sort | Lei Zhang |
| collection | DOAJ |
| description | Network alignment aims to identify the correspondence of nodes between two or more networks. It is the cornerstone of many network mining tasks, such as cross-platform recommendation and cross-network data aggregation. Recently, with the development of network representation learning techniques, researchers have proposed many embedding-based network alignment methods. The effect is better than traditional methods. However, several issues and challenges remain for network alignment tasks, such as lack of labeled data, mapping across network embedding spaces, and computational efficiency. Based on the graph neural network (GNN), we propose the URNA (unsupervised rapid network alignment) framework to achieve an effective balance between accuracy and efficiency. There are two phases: model training and network alignment. We exploit coarse networks to accelerate the training of GNN after first compressing the original networks into small networks. We also use parameter sharing to guarantee the consistency of embedding spaces and an unsupervised loss function to update the parameters. In the network alignment phase, we first use a once-pass forward propagation to learn node embeddings of original networks, and then we use multi-order embeddings from the outputs of all convolutional layers to calculate the similarity of nodes between the two networks via vector inner product for alignment. Experimental results on real-world datasets show that the proposed method can significantly reduce running time and memory requirements while guaranteeing alignment performance. |
| first_indexed | 2024-03-11T09:35:19Z |
| format | Article |
| id | doaj.art-30bff0664a7f4179b88e949246b9a240 |
| institution | Directory Open Access Journal |
| issn | 2227-7390 |
| language | English |
| last_indexed | 2024-03-11T09:35:19Z |
| publishDate | 2023-01-01 |
| publisher | MDPI AG |
| record_format | Article |
| series | Mathematics |
| spelling | doaj.art-30bff0664a7f4179b88e949246b9a2402023-11-16T17:21:31ZengMDPI AGMathematics2227-73902023-01-0111357310.3390/math11030573An Unsupervised Rapid Network Alignment Framework via Network CoarseningLei Zhang0Feng Qian1Jie Chen2Shu Zhao3School of Mathematics and Computer Science, Tongling University, Tongling 244061, ChinaSchool of Mathematics and Computer Science, Tongling University, Tongling 244061, ChinaSchool of Computer Science and Technology, Anhui University, Hefei 230601, ChinaSchool of Computer Science and Technology, Anhui University, Hefei 230601, ChinaNetwork alignment aims to identify the correspondence of nodes between two or more networks. It is the cornerstone of many network mining tasks, such as cross-platform recommendation and cross-network data aggregation. Recently, with the development of network representation learning techniques, researchers have proposed many embedding-based network alignment methods. The effect is better than traditional methods. However, several issues and challenges remain for network alignment tasks, such as lack of labeled data, mapping across network embedding spaces, and computational efficiency. Based on the graph neural network (GNN), we propose the URNA (unsupervised rapid network alignment) framework to achieve an effective balance between accuracy and efficiency. There are two phases: model training and network alignment. We exploit coarse networks to accelerate the training of GNN after first compressing the original networks into small networks. We also use parameter sharing to guarantee the consistency of embedding spaces and an unsupervised loss function to update the parameters. In the network alignment phase, we first use a once-pass forward propagation to learn node embeddings of original networks, and then we use multi-order embeddings from the outputs of all convolutional layers to calculate the similarity of nodes between the two networks via vector inner product for alignment. Experimental results on real-world datasets show that the proposed method can significantly reduce running time and memory requirements while guaranteeing alignment performance.https://www.mdpi.com/2227-7390/11/3/573network representation learningnetwork alignmentgraph neural networknetwork coarseningmulti-level embedding |
| spellingShingle | Lei Zhang Feng Qian Jie Chen Shu Zhao An Unsupervised Rapid Network Alignment Framework via Network Coarsening Mathematics network representation learning network alignment graph neural network network coarsening multi-level embedding |
| title | An Unsupervised Rapid Network Alignment Framework via Network Coarsening |
| title_full | An Unsupervised Rapid Network Alignment Framework via Network Coarsening |
| title_fullStr | An Unsupervised Rapid Network Alignment Framework via Network Coarsening |
| title_full_unstemmed | An Unsupervised Rapid Network Alignment Framework via Network Coarsening |
| title_short | An Unsupervised Rapid Network Alignment Framework via Network Coarsening |
| title_sort | unsupervised rapid network alignment framework via network coarsening |
| topic | network representation learning network alignment graph neural network network coarsening multi-level embedding |
| url | https://www.mdpi.com/2227-7390/11/3/573 |
| work_keys_str_mv | AT leizhang anunsupervisedrapidnetworkalignmentframeworkvianetworkcoarsening AT fengqian anunsupervisedrapidnetworkalignmentframeworkvianetworkcoarsening AT jiechen anunsupervisedrapidnetworkalignmentframeworkvianetworkcoarsening AT shuzhao anunsupervisedrapidnetworkalignmentframeworkvianetworkcoarsening AT leizhang unsupervisedrapidnetworkalignmentframeworkvianetworkcoarsening AT fengqian unsupervisedrapidnetworkalignmentframeworkvianetworkcoarsening AT jiechen unsupervisedrapidnetworkalignmentframeworkvianetworkcoarsening AT shuzhao unsupervisedrapidnetworkalignmentframeworkvianetworkcoarsening |