Privacy Preservation Algorithms Of Uncertain Graphs For Social Networks Base On Triadic Closure

In recent years,the development of various social software and network communication science are in a state of rapid progress.China's focus on and investment in 5G makes the construction of 5G infrastructure in China is developing rapidly.5G network has the characteristics of faster information transmission rate,shorter network response delay,larger network transmission bandwidth,more connected terminal devices and a larger amount of data exchanged.The arrival of 5G is also bringing more people to the Internet,allowing social networks to store vast amounts of user data.Therefore,the social network for data analysis and data mining to more scientific than traditional data analysis method is better,but the conference was carried out on the social network brings the problems such as personal privacy,so,for the purpose of data mining,scientific research,marketing,decision support,etc.)before release data need to be anonymous,used to protect user privacy in social networking from the attacker.Firstly,this thesis proposes the concept of latent edge based on the special nature of ternary closure principle,aiming at the characteristics of dynamic changes of social privacy.Ternary closure is a natural mechanism to establish connections between users in social networks.The network growth model of ternary closure can make the social network community appear naturally.If two people in the same social network graph have a mutual friend,there are potential edges between those two people.By adding potential edges to the original social network graph,the changes of the social network can be predicted to a certain extent,and for the social network graph under different conditions,we can achieve different degrees of privacy protection effect by adjusting the number of potential edges added.Then,based on today's social network protection method of the disadvantage of poor figure data of utility,social network is put forward to figure into uncertain figure method,to achieve the effect of modified figure attributes of social network diagram,reached in at the same time to protect the privacy of social network diagram that generate new data utility purpose.The main method is to inject uncertainty into some edge sets of the social network graph after adding potential edges,so that the degree of graph nodes and the existence of edges of the new graph will be changed,and the attacker will find the target of attack accurately,so as to achieve the effect of privacy protection for the social network graph.Because of the character of uncertain graph,this method has less loss of graph utility after modifying the original graph.Finally,due to the special properties of uncertain social network graph,in order to measure the attributes of nodes in uncertain social network graph,this thesis proposes a calculation method for the degree of nodes in uncertain graph corresponding to the concept of node degree in certain graph,and uses this method to measure the data utility of generated uncertain social network graph.In addition,in order to measure the privacy protection effect of the algorithm on the social network graph,this thesis uses the special concept of edge entropy to measure the privacy protection ability of the algorithm.Experimental results show that the proposed algorithm has less damage to graph utility when it has a certain degree of privacy protection ability,and it also has better protection effect on dynamic social networks because of the introduction of potential edges.