Institution profile

Center for Systems Biology Dresden

Academic institutioneurope · de
Official website
Research library5linked papers
Opportunities0open roles
Selected work

Representative Papers

Ollivier's Ricci Curvature on Complex-weighted Graphs

Aug 10, 2026

Existing discrete Ricci curvature methods struggle to handle networks with directionality and complex-valued weights, limiting their applicability in domains such as social, biological, and quantum systems. This work presents the first extension of Ollivier–Ricci curvature to complex-weighted graphs—encompassing directed graphs as a special case—and establishes a theoretical connection between this curvature and the magnetic Laplacian. By leveraging local neighborhood cycle structures, we derive rigorous upper and lower bounds for the curvature. Integrating optimal transport theory, combinatorial graph theory, and numerical optimization, we introduce the first well-defined Ollivier curvature for complex-weighted graphs, thereby unifying the treatment of directed edges and complex weights. The proposed curvature estimation algorithm demonstrates strong empirical performance and practical utility in community detection tasks on directed networks.

0 citationsRead paper
Recent publications

Latest Papers

Ollivier's Ricci Curvature on Complex-weighted Graphs

Aug 10, 2026

Existing discrete Ricci curvature methods struggle to handle networks with directionality and complex-valued weights, limiting their applicability in domains such as social, biological, and quantum systems. This work presents the first extension of Ollivier–Ricci curvature to complex-weighted graphs—encompassing directed graphs as a special case—and establishes a theoretical connection between this curvature and the magnetic Laplacian. By leveraging local neighborhood cycle structures, we derive rigorous upper and lower bounds for the curvature. Integrating optimal transport theory, combinatorial graph theory, and numerical optimization, we introduce the first well-defined Ollivier curvature for complex-weighted graphs, thereby unifying the treatment of directed edges and complex weights. The proposed curvature estimation algorithm demonstrates strong empirical performance and practical utility in community detection tasks on directed networks.

0 citationsRead paper