Investigating the Effectiveness of Laplacian-Based Kernels in Hub Reduction