Abstract This study investigates whether engineered relational database schemas exhibit small-world network structure, a topological property widely observed in natural and complex systems. We model relational schemas as graphs in which tables are nodes and foreign key relationships are edges, and analyze eight real-world open-source database systems spanning enterprise resource planning, workflow orchestration, healthcare, education, collaboration, and analytics domains. For each schema, we compute clustering coefficient, average shortest path length, and the small-world coefficient (sigma) relative to equivalent Erdős–Rényi random graph baselines. Results show that seven of eight schemas demonstrate statistically significant small-world organization, with sigma values ranging from 3.08 to 56.13. These findings suggest that efficient relational schema design may naturally converge toward small-world topologies that balance modular clustering with short global connectivity. The results support the broader hypothesis that computational architectures may benefit from topology-aware design principles inspired by complex network theory.
Jinhua Liao (Sun,) studied this question.