Nowadays, identification and detection community structures in complexnetworks is an important factor in extracting useful information from networks.Label propagation algorithm with near linear-time complexity is one of the mostpopular methods for detecting community structures, yet its uncertainty andrandomness is a defective factor. Merging LPA with other community detectionmetrics would improve its accuracy and reduce instability of LPA. Consideringthis point, in this paper we tried to use edge betweenness centrality toimprove LPA performance. On the other hand, calculating edge betweennesscentrality is expensive, so as an alternative metric, we try to use local edgebetweenness and present LPA-LEB (Label Propagation Algorithm Local EdgeBetweenness). Experimental results on both real-world and benchmark networksshow that LPA-LEB possesses higher accuracy and stability than LPA whendetecting community structures in networks.
展开▼
