A0644
Title: High order joint embedding for multi level link prediction
Authors: Yubai Yuan - Penn State University (United States) [presenting]
Abstract: Link prediction infers potential links from observed networks and is one of the essential problems in network analyses. In contrast to traditional graph representation modelling, which only predicts two-way pairwise relations, a novel tensor-based joint network embedding approach is proposed on simultaneously encoding pairwise links and hyperlinks onto a latent space, which captures the dependency between pairwise and multi-way links in inferring potential unobserved hyperlinks. The major advantage of the proposed embedding procedure is that it incorporates both the pairwise relationships and subgroup-wise structure among nodes to capture richer network information. In addition, the proposed method introduces a hierarchical dependency among links to infer potential hyperlinks and leads to better link prediction. Theoretically, the estimation consistency is established for the proposed embedding approach and provides a faster convergence rate than link prediction utilizing pairwise links or hyperlinks only. Numerical studies on both simulation settings and Facebook ego networks indicate that the proposed method improves both a hyperlink and pairwise link prediction accuracy compared to existing link prediction algorithms.
