The discovery of community structures in large-scale, complex networks is of fundamental importance to social network analysis. This task has enormous implications for understanding information flow and the behavioral characteristics of groups. This paper introduces a new hybrid model that combines the DBSCAN algorithm with a modified Cuckoo Search Optimization (CSO)algorithm of L´evy flight. This synergy leverages DBSCAN's robust capability to isolate noise and identify dense core nodes, while the optimized CSO performs a global search for the most effective community partitions. The proposed model was rigorously evaluated against a suite of established algorithms, including the Bat Algorithm, AFSA, Multilevel, Walktrap, AKHSO, Ant-Lion Optimizer, Lion Optimization Algorithm, and standard Cuckoo Search. Experimental values of four standard social networks substantiate the better performance of our method, which obtained the top recorded results for both Modularity and NMI. These findings indeed validate that the approach not only detects communities with higher internal cohesion but also more properly mirrors the observed ground-truth structures, verifying its effectiveness and strength.
Babers, R. (2025). DBSCAN with Cuckoo Search Algorithm for Social Networks Community Detection. Advances in Basic and Applied Sciences, 6(1), 19-25. doi: 10.21608/abas.2025.415766.1081
MLA
Ramadan Babers. "DBSCAN with Cuckoo Search Algorithm for Social Networks Community Detection", Advances in Basic and Applied Sciences, 6, 1, 2025, 19-25. doi: 10.21608/abas.2025.415766.1081
HARVARD
Babers, R. (2025). 'DBSCAN with Cuckoo Search Algorithm for Social Networks Community Detection', Advances in Basic and Applied Sciences, 6(1), pp. 19-25. doi: 10.21608/abas.2025.415766.1081
VANCOUVER
Babers, R. DBSCAN with Cuckoo Search Algorithm for Social Networks Community Detection. Advances in Basic and Applied Sciences, 2025; 6(1): 19-25. doi: 10.21608/abas.2025.415766.1081
)
View on SCiNiTO