Link Prediction In Social Networks Using Hyper-Motif Representation On Hypergraph

Link prediction, a critical pursuit in complex networks research, revolves around the predictive understanding of connections between nodes. Our novel approach introduces a hypergraph to model the network, diverging from the conventional “node–edge” structures. This departure involves the strategic mapping of hyper-motifs into open and closed structures, and the utilization of hyper-motifs as hyper-nodes. This paper proposes the learning embedding based on hyper-motif of the network (LEHMN) model to improve the link prediction process. This innovative framework aims at enhancing our capacity to discern and represent nuanced structural similarities between nodes that might elude traditional models. To further refine our approach, we introduce the depth and breadth motif random walk strategy. This strategy, designed with a consideration for both connectivity and structural similarity, excels in acquiring node sequences. We apply this method to both open and closed hyper-motif structures, emphasizing its versatility and effectiveness in capturing the intricate relationships inherent in the network. In the realm of experimental validation, our proposed model outshines state-of-the-art baselines when tested on diverse datasets. The results highlight significant improvements across various scenarios, underscoring the robustness and efficacy of our approach. This substantiates the pivotal role our method plays in advancing link prediction within the dynamic landscape of complex networks research.