A New Link Prediction Algorithm: Node Link Strength Algorithm

Link prediction is an important content in the related fields of social networks. Nowadays, the link prediction algorithm based on node similarity is the research hotspot. Node similarity is usually calculated by common neighbours, however, the impact of each common neighbour on the link is different, and how to distinguish the different roles of common neighbours needs to be further studied. The paper proposes the concept of node link strength and redefines the calculation method of node similarity. The new algorithm takes into account the link strength between the common neighbour nodes and the predicted nodes, better to reflect the effect of each common neighbour node. The algorithm is carried out experiments on five real social network datasets, the prediction results show that the algorithm´s prediction accuracy is more effective than the existing mainstream algorithms based on node similarity and maintains low time complexity.