The Trust Value Calculating for Social Network Based on Machine Learning

In this paper, a social network model is built for the social network information in the social network, and the machine learning method is used to calculate the node trust value. First, the results calculated by the traditional node trust value calculation method and some auxiliary information are used as the training feature of the machine learning, and the measurement whether there is edge between nodes as label information. Second, the node logistic regression model is used as the training model to calculate the node trust value. Then, recommendation algorithm which is analogous to the user collaborative filtering algorithm is used to calculate node trust value. At last, the simulation is used to verify the performance of the improved method, and the results show that the prediction accuracy of node trust value computing by improved algorithm is significantly higher than that of node trust value computing by formula.