Directed mixed membership stochastic blockmodel

Abstract:Mixed membership modeling for undirected networks has been extensively explored in network science over the past few years. Despite the substantial progress made for undirected cases, handling mixed membership structures in directed networks continues to pose substantial difficulties. To address this gap, we introduce the Directed Mixed Membership Stochastic Blockmodel (DiMMSB), a novel framework tailored for directed networks with overlapping communities. A key feature of DiMMSB is its ability to treat the row and column nodes of the adjacency matrix as distinct entities, each potentially following its own community organization. Building on this model, we develop DiSP, an efficient spectral procedure to estimate mixed memberships for both sets of nodes. Through delicate analysis, we derive node-specific error bounds of DiSP under mild sparsity conditions. Simulation results support the theoretical results, demonstrating that DiSP achieves lower error rates and faster computation than its competitor. Moreover, applications to real data highlight DiSP's effectiveness in uncovering asymmetric structural patterns.

Machine Learning (stat.ML); Machine Learning (cs.LG); Social and Information Networks (cs.SI)

arXiv:2101.02307 [stat.ML]

(or arXiv:2101.02307v4 [stat.ML] for this version)

https://doi.org/10.48550/arXiv.2101.02307 Focus to learn more arXiv-issued DOI via DataCite

Qing H, Wang J. Directed mixed membership stochastic blockmodel. Information Sciences. 2026 Apr 30:123577

https://doi.org/10.1016/j.ins.2026.123577 Focus to learn more DOI(s) linking to related resources

arXivLabs is a framework that allows collaborators to develop and share new arXiv features directly on our website.

Both individuals and organizations that work with arXivLabs have embraced and accepted our values of openness, community, excellence, and user data privacy. arXiv is committed to these values and only works with partners that adhere to them.

Have an idea for a project that will add value for arXiv's community? Learn more about arXivLabs.