Link prediction algorithm for multiplex networks based on the intermediary probability model

Link prediction is an important research direction in complex network analysis as it can discover and predict missing links in networks as well as potential new links in the future, aiding in the study of network evolution patterns. Previous work has shown that there is a certain correlation between the layers of a multiplex network, and leveraging information from other layers can effectively improve the performance of link prediction in the target layer. However, effectively extracting and utilizing information from different layers in multiplex networks remains a challenging task. To address the issue of link prediction in multiplex networks, this paper proposes an algorithm for multiplex network link prediction based on the intermediary probability model. This paper focuses on triangular relationships between node pairs and their neighbors in multiplex networks, considering both intralayer and interlayer relationships. It proposes a method to calculate cross-layer clustering coefficients, measuring the contribution of common neighbors in other layers to link formation in the target layer. Then, the intermediary probability model integrates these contributions, resulting in the link formation probability. This approach outperforms other link prediction methods for multiplex networks, as verified through experiments on nine real-world datasets.