DomiCom: Discovery of top-k dominant communities in networks with node attributes
Georgiadis, Nikolaos Tiakas, Eleftherios Papadopoulos, Apostolos
Georgiadis, Nikolaos
Tiakas, Eleftherios
Papadopoulos, Apostolos
Supervisor
Department
Computer Science
Embargo End Date
Type
Journal article
Date
License
Language
English
Collections
Research Projects
Organizational Units
Journal Issue
Abstract
The community structure is an inherent property of real-world networks. Broadly, a set of nodes S forms a community if its nodes exhibit a significantly higher level of interconnection with each other than with nodes outside S. The precise notion of a community, however, depends on the mathematical formulation adopted. The problem becomes more intriguing when graph nodes are enriched with attributes, as these attributes can influence how communities are defined. In this work, we incorporate node attributes to construct more meaningful communities that reflect both structural connections and attribute information. Specifically, we introduce the concept of dominance relationships between nodes: a node u is considered more important than a node v if the attributes of u dominate those of v. Experimental evaluation on real-world attributed networks demonstrate the efficiency and effectiveness of the proposed approach. The reported communities are meaningful and robust based on their significance and structural coherence.
Citation
N. Georgiadis, E. Tiakas, A. Papadopoulos, "DomiCom: Discovery of top-k dominant communities in networks with node attributes," South African Computer Journal, vol. 23, no. 00, pp. 73-73, 2026, https://doi.org/10.2298/csis240430073g.
Source
South African Computer Journal
Conference
Keywords
46 Information and Computing Sciences, 4605 Data Management and Data Science
Subjects
Source
Publisher
National Library of Serbia