Research On Valish Link Prediction In Social Networks

In recent years, social network, as one of the main popular network applications all over the world, has become an important way to offer people a chance to share information and communicate with each other. This paper aims to probe into the link prediction problem in social networks. The establishment of links in link prediction has been studied for several years and has formed rather mature methods. However, fewer than a dozen people are looking in this direction. Therefore, this paper proposes two kinds of solution to predict the valish links in social networks. One is based on computing the strength of links, and the other is based on subgraph transition in social networks. The primary study in this paper are described as the following parts:(1) Constructs the social networks based on the datasets. Analyses the features of valish links in social networks and figure out the reason why links disappear or exist.(2) Proposes a solution for valish link prediction based on computing the strength of links in undirected social networks. Develops measures of valish link prediction respectively on networks without weight, networks with weight, and sequential networks. Implement these measures to test and compare the results of prediction. Extend valish link prediction based on link strength to directed social networks. Develops measures of valish link prediction respectively on directed networks without weight, networks with weight, and sequential networks. Implement these measures to test and compare the results of prediction.(3) Analyses the law of subgraph transition in social networks, and figure out the probability matrix of subgraph transition STM. Design new valish prediction algorithm based on STM. Implement the designed process to test and verify the advantages of the prediction algorithm.