Research On K Anonymity Algorithm Based On Social Network

With the rapid development of Internet technology,various social networks have generated a large amount of user data,and the way of publicly releasing social network data is often accompanied by the risk of leakage of users’ personal privacy information.Anonymization is one of the main technologies to solve the problem of privacy leakage in data release,but most anonymization methods still have huge information loss and large disturbance to the community structure in the original social network.Excessive anonymization of attributes leads to low utility of anonymized data.In view of the above problems,this paper conducts the following research.Firstly,in view of the problem that the existing K degree anonymity algorithm cannot meet the needs of community structure analysis and the large degree of information loss in terms of data utility,the label entropy is introduced to comprehensively consider the characteristics and global characteristics of the nodes,and a K degree anonymity algorithm based on community division is proposed.This method calculates the label entropy of each node,divides the network node into several communities by using the label propagation algorithm,partitions the degree sequence of the graph and operates K anonymity,and modifies the graph by combining the community where the node is located and the neighborhood correlation of the node,so that the original social network achieves K anonymity.Secondly,in view of the problem that the attributes are too anonymized and the graph is modified too much,the anonymized data is not very practical.The concept of neighborhood attribute similarity is introduced,and an attribute anonymization algorithm based on node similarity is proposed.This method is based on the greedy strategy,and the degree sequence is grouped by calculating the similarity of the attributes of the neighborhood.At the same time of grouping,it is judged whether there are L kinds of attribute values of the sensitive nodes.The attribute modification operation makes the original social network graph satisfy the sensitive attribute L diversity anonymity while achieving K anonymity.Finally,the above algorithms are compared with the existing algorithms on different real data sets,and the existing algorithms are compared and analyzed.