An Algorithm Based on Non-Negative Matrix Factorization for Detecting Communities in Networks

Chenze Huang,Ying Zhong

DOI: https://doi.org/10.3390/math12040619

IF: 2.4

2024-02-20

Mathematics

Abstract:Community structure is a significant characteristic of complex networks, and community detection has valuable applications in network structure analysis. Non-negative matrix factorization (NMF) is a key set of algorithms used to solve the community detection issue. Nevertheless, the localization of feature vectors in the adjacency matrix, which represents the characteristics of complex network structures, frequently leads to the failure of NMF-based approaches when the data matrix has a low density. This paper presents a novel algorithm for detecting sparse network communities using non-negative matrix factorization (NMF). The algorithm utilizes local feature vectors to represent the original network topological features and learns regularization matrices. The resulting feature matrices effectively reveal the global structure of the data matrix, demonstrating enhanced feature expression capabilities. The regularized data matrix resolves the issue of localized feature vectors caused by sparsity or noise, in contrast to the adjacency matrix. The approach has superior accuracy in detecting community structures compared to standard NMF-based community detection algorithms, as evidenced by experimental findings on both simulated and real-world networks.

mathematics

What problem does this paper attempt to address?

The problem that this paper attempts to solve is the accuracy problem of detecting community structure in sparse networks. Specifically, the existing community detection algorithms based on Non - Negative Matrix Factorization (NMF) experience performance degradation when dealing with data matrices of low density due to the localization of eigenvectors. This localization phenomenon usually occurs in sparse or noisy networks, making it difficult for the NMF method to effectively reveal the global structure of the network.

To overcome this challenge, the paper proposes a new NMF - based community detection algorithm, named CDNMF (Community Detection using Non - negative Matrix Factorization). This algorithm effectively solves the eigenvector localization problem by using local eigenvectors to represent the topological features of the original network and learning the regularization matrix. Through this method, CDNMF can not only better capture the global structure of the data matrix but also achieve more accurate community detection in sparse networks.

The main contributions of the paper include:
1. Proposing a matrix regularization process to enhance the representation ability of the overall topological features of the network while solving the problem of data matrix eigenvector localization.
2. Designing a method for discovering sparse network communities by decomposing non - negative matrices. This method combines regularization transformation and spectral analysis of non - backtracking matrices and can effectively determine the number of community divisions without increasing computational complexity.
3. Experimental results show that the proposed CDNMF algorithm is superior to the existing NMF - based community detection techniques in terms of the accuracy of community segmentation, especially in sparse networks.

Through these improvements, the paper provides a more efficient and accurate method for sparse network community detection, providing a powerful tool for complex network structure analysis.