Year: 2025
Type: conference-paper
Abstract: In the era of big data, particularly biological big data, integrating node attributes with topological structure is essential for understanding complex networks. This chapter introduces node attribute entropy (NAE), a method combining network topology and node attributes (e.g., gene expression) to quantify node importance. By integrating weight factors and local entropy, NAE reveals the nonlinear relationship between node weights and attribute entropy. Theoretical analysis highlights its properties under unconstrained and globally normalized conditions, with network structures (e.g., scale-free, small-world) significantly influencing entropy dynamics. Compared to traditional centrality metrics and other entropy measures, NAE excels in capturing node diversity and information efficiency. Experiments on a colon cancer gene co-expression network demonstrate NAE's superior accuracy in identifying key genes and biologically relevant modules. This work provides a robust framework for complex network analysis, with applications in biological and social networks.
Authors: Qingcai He, Binghui Guo, Zhiming Zheng
Institution: Beihang University
Language: English
Subfield: Statistical and Nonlinear Physics
Field: Physics and Astronomy
Domain: Physical Sciences
Open Access status: hybrid
OpenAlex ID: W4413812424