Abstract:

To solve the privacy-preserving datapublishing problem in context of vertical distribution，the definition of global k-anonymity and its necessary and sufficient conditions are given based on the concept of equivalent groups of tuples, the global equivalent groups of tuples are solved by using a polynomial representation method, a distributed privacy preserving k-anonymity model which has the character of privacy, accuracy and fairness by applying the set polynomial indication method and homomorphic encryption protocol based on secure multi-party computation is proposed. The model can help each data owner to securely publish real and affect data set for statistical analyzing and data mining without revealing sensitive original information. Experiments demonstrate that this model can provide the good capability of security, accuracy and efficiency.