A network reconfiguration and overlapping community detection joint method, device, equipment and medium

By introducing community membership and responsibility variables into complex networks and using the expectation-maximization algorithm to iteratively optimize the community matrix, the problems of insufficient accuracy and poor noise robustness in complex network reconstruction in existing technologies are solved, and accurate detection of underlying overlapping community structures and efficient reconstruction of network topology are achieved.

CN121786595BActive Publication Date: 2026-06-02SHENZHEN UNIV
View PDF 2 Cites 0 Cited by

Patent Information

Authority / Receiving Office
CN · China
Patent Type
Patents(China)
Current Assignee / Owner
SHENZHEN UNIV
Filing Date
2026-03-03
Publication Date
2026-06-02

AI Technical Summary

Technical Problem

Existing technologies cannot fully utilize the characteristic traces of community structure in binary state evolution, resulting in insufficient accuracy in complex network reconstruction and a lack of robustness to noise and errors. It is difficult to accurately detect the underlying overlapping community structure and reconstruct the hidden network topology from extremely limited binary state time series observation data.

Method used

By acquiring binary time-series data and the number of communities in the community network, we initialize the community membership matrix and interaction matrix, iteratively optimize the community membership matrix and interaction matrix using the expectation-maximization algorithm, introduce community membership variables and responsibility variables, track the contribution of each edge to the observation dynamics, explicitly model the community structure and noise term, and form a unified optimization architecture.

Benefits of technology

It enables the detection of underlying overlapping community structures in complex networks and accurate reconstruction of hidden network topologies, improving reconstruction accuracy, reducing dependence on the quality of observation data, enhancing fault tolerance to noise, and simplifying model parameter tuning.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN121786595B_ABST
    Figure CN121786595B_ABST
Patent Text Reader

Abstract

The application discloses a network reconstruction and overlapping community detection joint method, device, equipment and medium, the method comprises the following steps: obtaining binary time series data and community quantity of community network; initializing community membership matrix and community interaction matrix; using expectation maximization algorithm to iterate community membership matrix and community interaction matrix to obtain target community membership matrix; determining the community structure of the community network and the overlapping community in the community network based on the target community membership matrix. The application explicitly introduces the community membership variable in the parameter space, so that the network inference and community detection, which are regarded as independent problems, form a unified optimization architecture. The responsibility variable is used to track the specific contribution of each edge to the observed dynamics, ensuring the deep coupling and mutual promotion of network inference and community detection. The application not only realizes the detection of the underlying overlapping community structure of the complex network and the accurate reconstruction of the hidden network topology structure, but also ensures the reconstruction accuracy of the complex network.
Need to check novelty before this filing date? Find Prior Art