Font Size: a A A

A Personalized Data Anonymous Publishing Model Based On Feedback Penalty Mechanism

Posted on:2018-06-23Degree:MasterType:Thesis
Country:ChinaCandidate:Z H ZhengFull Text:PDF
GTID:2348330518954371Subject:Network and optical communication
Abstract/Summary:PDF Full Text Request
As the most common personal data release scenario in this information age,the web services based on personal information has inevitably became the hardest hit.The privacy protection is provided only by communication network system and business services in the traditional network service,but when the users' information become the treasure,the users' personal information was collected maliciously,and the users become "no secret person" and will be attacked at an unknown time by someone.From privacy protection perspective,this paper discusses how to balance the privacy information protection and its usefulness in interactive personal data publishing scenario.After a comprehensive analysis about all kinds of anonymous protection principles,a new idea that introducing feedback punishment mechanism in the framework of traditional stochastic game theory,and using the sum of individual attribute disclosure risk is less than personal privacy tolerance or not to correct the error of the game results is put out,and the results of an experiment verified the personalized data anonymity released model based on feedback punish mechanism.By solving the mixed strategy Nash equilibrium of the game,which is based on the service process between servers and users could be abstract as a mixed strategy complete information static game,the users can choose the best strategy to get maximum income.Our experiment proved that the model was effective on developing the former,the conclusion is mainly reflected in two points: 1)The proposed model has stable data utility rate,privacy protection degree,and contribution rate,which will not change with the increase of the times of service.2)The fact that the effect of this model was determined by personalized property configuration itself,in other words,different users different data utility rate,privacy protection degree,which shows the personalization of this model,and ensure the balance of maximum utility between data privacy protection.
Keywords/Search Tags:feedback punishment, game theory, interactive scenario, anonymous publishing, personalization
PDF Full Text Request
Related items