OGDOS: An Open Graph Dataset Organized by Scales

Graph-related technologies, including social networks, transportation systems, and bioinformatics, are continually evolving in various application domains. The advancement of these technologies often relies on high-quality graph datasets for validating scalability, complexity, and performance. However, existing datasets are typically categorized by domain or type and lack explicit organization at the scale level. Moreover, existing datasets may not fully cover a wide range of graph scale levels. These absences may limit comprehensive evaluations of the complex relationships between graph scales and performance.

To address these gaps, we introduce an open graph dataset organized by scales named OGDOS. Our OGDOS is an open graph dataset organized by scales, encompassing 470 preset scale levels, with node counts ranging from 100 to 200,000 and edge-to-node ratios from 1 to 10. The dataset combines scale-aligned real-world graphs and synthetic graphs, providing a versatile resource for evaluating various graph-related applications. Each graph in OGDOS is stored as a “.json” file in a specific directory path, which reflects its corresponding scale levels. For example, a “.json” file located at the “…\OGDOS \ 100 \ 1 \” folder represents a graph with a node scale of 100 and an edge-to-node ratio of 1. The source types of these graphs in OGDOS are provided in Table 2.