MINING: Multi-Granularity Network Alignment Based on Contrastive Learning
Zhongbao Zhang, Shuai Gao, Sen Su, Li Sun, Ruiyang Chen
Abstract
Network alignment aims to discover nodes in different networks belonging to the same identity. In recent years, the network alignment problem has aroused significant attentions in both industry and academia. However, the continuous exploding of network data brings two challenges in solving the network alignment problem, i.e., large network scale and scarce labeled data. To bridge this gap, in this paper we propose a novel approach termed as <u>M</u> ulti-granular <u>I</u> ty <u>N</u> etwork al <u>I</u> gnment based on co <u>N</u> trastive learnin <u>G</u> (MINING). Specifically, in MINING, we first design multi-granularity alignment framework to solve the issue of large network scale. Then, we design intra- and inter-network contrastive learning to solve the issue of scarce labeled data. Moreover, we provide theoretical proofs to demonstrate the effectiveness of MINING. Finally, we conduct extensive experiments on the benchmark datasets of Facebook-Twitter, AMiner-LinkedIn and DBpedia <inline-formula><tex-math notation="LaTeX">$_{\text{ZH}}$</tex-math></inline-formula> -DBpedia <inline-formula><tex-math notation="LaTeX">$_{\text{EN}}$</tex-math></inline-formula> , and results show that MINING can averagely achieve 15.93% higher <inline-formula><tex-math notation="LaTeX">$\operatorname{Hits@}k$</tex-math></inline-formula> and 14.82% higher <inline-formula><tex-math notation="LaTeX">$\operatorname{MRR@}k$</tex-math></inline-formula> compared with the state-of-the-art methods.