Relational Data Privacy Protection In Data Exchange

Data exchange,as a type of emerging industry,breaks the information island and the industry information barriers to achieve the maximization of the value of the data.In the process of data exchange,the data is often released in the form of free publication or exchange,while the released data contains a large number of personal information.According to the personal information,the attacker can use the data mining technology to analyze information,so as to make the personal privacy be leaked,and lead to economic loss and personal injury and so on.Thus,how to effectively protect the privacy of data exchange is a key issue to be solved urgently.To ensure the security of privacy,we need to do full autonomy processing for the individual privacy,which is a reflection of ownership of personal data,and is also the effective and active means of protecting personal privacy.At the same time,since real-time data is very valuable,how to prevent the privacy of real-time data dissemination and ensure the security of data exchanges is also an important issue that should be solved urgently.At present,researchers have proposed lots of privacy protection methods,but these models cannot effectively protect the privacy of fully autonomous data,and there are not reasonable privacy protection models on privacy protection of the real-time release.In order to protect the security of the privacy of the fully autonomous data,this dissertation proposes an individual (?,?)-anonymity model.In the model,a and w respectively represent the constraint value of thesensitive attributes(S)and the QI attributes,which both are set by providers.The(?,?)-anonymity model is combined with the granular computing and the top-down local recoding to process the datasets of the providers.The core of the model is to meet all individuals' requirements.Moreover,the performance analysis shows that this model not only satisfies the individualized privacy requirements,but also brings higher efficiency and lower information loss.Meanwhile,in order to ensure the privacy of the data in real-time release,by applying the fuzzy processing and the m-signature constraint which is proposed by the author,this dissertation constructs a new privacy protection model for the real-time release.Finally,we use experiments to demonstrate the performance of the proposed models.The models proposed in this dissertation not only protect the privacy of data exchange,but also lay a practical foundation for the data exchange.What's more,it will further promote the diversification of markets of data exchange,then maximize the value of the personal data and promote the development of innovation of data industry.