Abstract:We propose two methods for the unsupervised detection of communities in undirected multiplex networks. These networks consist of multiple layers that record different relationships between the same entities or incorporate data from different sources. Both methods are formulated as gradient flows of suitable energy functionals: the first (MPBTV) builds on the minimization of a balanced total variation functional, which we show to be equivalent to multiplex modularity maximization, while the second (DGFM3) directly maximizes multiplex modularity. The resulting non-linear matrix-valued ordinary differential equations (ODEs) are solved efficiently by a graph Merriman--Bence--Osher (MBO) scheme. Key to the efficiency is the approximate integration of the discrete linear differential operators by truncated eigendecompositions in the matrix exponential function. Numerical experiments on several real-world multiplex networks show that our methods are competitive with the state of the art with respect to various metrics. Their major benefit is a significant reduction of computational complexity leading to runtimes that are orders of magnitude faster for large multiplex networks.

What problem does this paper attempt to address?

### Problems Addressed by the Paper The paper aims to address the problem of community detection in multilayer networks. Specifically: 1. **Community Detection Methods**: The paper proposes two gradient flow-based methods for community detection in undirected multilayer networks. These methods achieve community detection by optimizing specific energy functionals. 2. **Definition of Multilayer Networks**: A multilayer network consists of multiple layers, each recording different relationships between the same entities or information from different data sources. There are inter-layer connections, but they are limited to connections between different layer instances representing the same entity. 3. **Objective Function for Community Detection**: The first method (MPBTV) achieves multilayer motif maximization by minimizing a balanced total variation functional; the second method (DGFM3) directly maximizes multilayer motifs. 4. **Algorithm Efficiency**: The proposed nonlinear matrix ordinary differential equations are efficiently solved using the graph Merriman–Bence–Osher (MBO) scheme and by approximating discrete linear differential operators with truncated eigenvalue decomposition. Numerical experiments show that these two methods run several orders of magnitude faster than existing methods when handling large-scale multilayer networks. 5. **Relation to Single-layer Networks**: The study extends existing techniques from single-layer networks to multilayer networks and demonstrates the equivalence between multilayer motif maximization and balanced total variation minimization. In summary, the paper proposes a novel approach to solving the problem of community detection in multilayer networks and achieves this goal through efficient algorithms, showing significant advantages in practical applications.

Gradient flow-based modularity maximization for community detection in multiplex networks

Discriminative community detection for multiplex networks

Community Detection in Multiplex Networks Based on Evolutionary Multitask Optimization and Evolutionary Clustering Ensemble

Community Detection in Multiplex Networks Using Orthogonal Non-Negative Matrix Tri-Factorization Based on Graph Regularization and Diversity

Community Detection and Improved Detectability in Multiplex Networks

Modularity functions maximization with nonnegative relaxation facilitates community detection in networks

Community Detection via Maximization of Modularity and Its Variants

Geometric Multiscale Community Detection: Markov Stability and Vector Partitioning

Detecting Composite Communities in Multiplex Networks: A Multilevel Memetic Algorithm

A Community Detection Algorithm of Multiplex Networks with Layer Reduction

Discrimination universally determines reconstruction of multiplex networks

Identifying modular flows on multilayer networks reveals highly overlapping organization in social systems

ModMRF: A Modularity-Based Markov Random Field Method for Community Detection

A New Multiobjective Evolutionary Algorithm for Community Detection in Dynamic Complex Networks

Modularity and community structure in networks.

A Graph Convolutional Fusion Model for Community Detection in Multiplex Networks

Multilayer flows in molecular networks identify biological modules in the human proteome

Optimizing Modularity with Nonnegative Matrix Factorization

Community Detection in Multiplex Networks by Deep Structure-Preserving Non-Negative Matrix Factorization

Multilink Communities of Multiplex Networks

Finding multifaceted communities in multiplex networks