Nonnegative Matrix Factorization Algorithm with Two Attribute Matrices for Community Detection

Community detection based on nonnegative matrix factorization (NMF) has the advantages of clear physical meaning, simple calculation and strong interpretability, but its accuracy needs to be improved. For this reason, this paper puts forward the community detection algorithm using NMF with two attri...

Ausführliche Beschreibung

Gespeichert in:
Bibliographische Detailangaben
Veröffentlicht in:2020 IEEE International Conference on Artificial Intelligence and Computer Applications (ICAICA) S. 87 - 90
Hauptverfasser: Zhao, Yingying, Xu, Hui, Zhou, Cheng
Format: Tagungsbericht
Sprache:Englisch
Veröffentlicht: IEEE 01.06.2020
Schlagworte:
Approximation algorithms
attribute information matrix
community detection
Detection algorithms
Linear programming
Matrix decomposition
Motion pictures
nonnegative matrix factorization algorithm
Partitioning algorithms
Symmetric matrices
Online-Zugang:Volltext
Tags: Tag hinzufügen
Keine Tags, Fügen Sie den ersten Tag hinzu!
Beschreibung
Zusammenfassung:Community detection based on nonnegative matrix factorization (NMF) has the advantages of clear physical meaning, simple calculation and strong interpretability, but its accuracy needs to be improved. For this reason, this paper puts forward the community detection algorithm using NMF with two attribute information matrices(2AMNMF). First of all, two attribute information matrices are created from calculating similarity between the entity and entity, then one of which is decomposed into two non-negative matrices by NMF, another attribute information matrix is added into objective function for optimization. Evaluation is made by modularity Q. The experiment results show that the algorithm of community detection we proposed is more accurate than the original NMF algorithm.
DOI:10.1109/ICAICA50127.2020.9182466