Search

Link prediction in bipartite networks is attracting tremendous research interests. Most previous studies mainly assume the generation of link follows a predefined prior mechanism while neglecting complexity of the link generation mechanisms. To address this limitation, we present a parameter-free method, termed S tructural E nhancement and S tructural P erturbation (SESP), which jointly exploits explicit relations (low-order information) and implicit relations (high-order information) from the perspective of perturbation. The essence of SESP is that it transforms bipartite link prediction into monopartite link prediction without losing any information and predicts the missing links from a perturbed perspective. Compared with traditional link prediction methods, SESP does not assume a particular link generation mechanism, but learns this mechanism from the network itself. Extensive experiments on several disparate real-world bipartite networks demonstrate the effectiveness of the SESP model. • We transform bipartite LP into the monopartite without losing structure information. • We propose SESP framework to model explicit and implicit relations simultaneously. • Experiments on eight real-world bipartite networks show the performance of SESP. [ABSTRACT FROM AUTHOR]

Link prediction aims to extract missing informations, identify spurious interactions and potential informations in complex networks. Similarity-based methods, maximum likelihood methods and probabilistic models are the mainstreaming classes algorithms for link prediction. Meanwhile, low rank matrix approximation has been widely used in networks analysis and it can extract more useful features hidden in the original data through some kernel-induced nonlinear mapping. In this paper, based on the non-negative matrix factorization (NMF), we propose a kernel framework for link prediction and network reconstruction by using different kernels which could get both global and local information of the network through kernel mapping. In detailed, we map the adjacency matrix of the network to another feature space by two kernel functions, the Linear Kernel and Covariance Kernel, which have the principled interpretations for the network analysis and link predication. We test the AUC and Precision of widely used methods on a series of real world networks with different proportions of the training sets, experimental results show that our proposed framework has more robust and accurate performance compared with state-of-the-art methods. Remarkably, our approach also has the potential to address the problem of link prediction using small fraction of training set. [ABSTRACT FROM AUTHOR]

Link prediction in bipartite networks aims to identify or predict possible links between nodes of different types based on known network information. However, most existing studies predominantly focus on monopartite networks, neglecting the intrinsic properties unique to bipartite networks, such as the intricate high-order relationships between nodes. Both explicit relations (representing low-order information) and implicit relations (representing high-order information) play essential roles in predicting the evolution of bipartite networks, and they are indispensable and mutually reinforce each other. To fully leverage their potential in addressing the link prediction problem, we propose a novel framework from the perspective of network representation. This framework not only effectively integrates explicit relations and implicit relations, but also preserves the local and global structure of bipartite networks. Specifically, the probability of a link between two nodes of different types is determined by the linear sum of the contribution values of the mutually connected nodes in their respective common neighbors. Implicit relations are then used to preserve the local structure during the network representation. Furthermore, we implement optimization using a relaxed majorization-minimization algorithm, offering the advantage of uncovering high-quality local minima. Our proposed framework has undergone extensive testing on eight real-world datasets, and the results unequivocally demonstrate its significant superiority over state-of-the-art methods. [ABSTRACT FROM AUTHOR]

• We introduce graph structural information learned from overlapped neighbors to maintain graph topology, thus assisting in link prediction. • We propose NOH that uses a heterogeneous hypergraph variational autoencoder to learn latent node embeddings. • We design a simple but effective structural information generator which can handle overlapped neighborhoods. • NOH consistently achieves state-of-the-art performance compared with the baselines on link prediction. In real world, a large number of networks are heterogeneous, containing different types of semantics and connections. Existing studies typically only consider lower-order pairwise relations rather than higher-order group interactions. Furthermore, they tend to focus more on node attributes rather than graph structural information. This results models failing to maintain graph topology effectively, which reduces the effectiveness on link prediction. To address these limitations, we propose N eighborhood O verlap-aware H eterogeneous hypergraph neural network (NOH) that learns useful structural information from the heterogeneous graph and estimates overlapped neighborhood for link prediction. Our model fuses the heterogeneity of graphs with structural information so that the model maintains both lower-order pairwise relations and higher-order complex semantics. Our extensive experiments on four real-world datasets show that NOH consistently achieves state-of-the-art performance on link prediction. [ABSTRACT FROM AUTHOR]