A Method of Link Prediction Based on Betweenness

Link prediction in complex networks has attracted increasing attention of researchers in many domains. The prediction methods are usually used to find missing information, identify spurious interactions, and reconstruct networks. Inspired by the rich-get-richer mechanism, we propose a novel index on the basis of betweenness. Extensive experiments show that the proposed method performs well on some networks. Especially, on the Adjnoun network and Florida network, it outperforms some mainstream link prediction baselines, such as CN Index, AA Index and RA Index.