Synthetic Signed Graphs

This dataset contains a collection of synthetic signed networks generated for benchmarking and evaluating community detection algorithms on signed graphs. The networks are available in both GEXF and GraphML formats.

The dataset generation process consists of two stages. First, unsigned graphs are generated using the Large Network Generator (see article "Large Network Generator: A simple, efficient, and flexible graph formation algorithm", DOI: 10.1088/2632-072X/ae6eab). Then, edge labels are introduced to transform the unsigned graphs into signed networks. The labeling process is guided by community structure in order to generate realistic signed topologies.