Graph Reduction in Multirelational Networks: A Spreading-Oriented Reduction Benchmark
Researchers have introduced the Spreading-Oriented Reduction Benchmark (SORB), an open-source framework designed to evaluate influence maximization (IM) models. SORB integrates graph reduction techniques directly into the evaluation process, allowing for a more comprehensive assessment of IM algorithms across various real-world network types and task settings. Initial studies using SORB indicate that the effectiveness of graph reduction is highly dependent on the network structure and the specific downstream task, with sparsification proving beneficial for single-layer networks but leading to degradation on flattened multilayer networks. AI