Loading...
| Friend's email | |
| Your name | |
| Your email | |
| enter code | |
This page was sent successfuly
- Type of Document: M.Sc. Thesis
- Language: Farsi
- Document No: 56920 (19)
- University: Sharif University of Technology
- Department: Computer Engineering
- Advisor(s): Jalili, Rasool
- Abstract:
- Usually, developers analyze data in order to enhance the services they provide to users, but this data analysis may compromise the privacy of the users during this process. Alternatively, service providers may choose to share their data with a third party for the purpose of data analysis, but they may have a strong aversion to sharing the personal information of their users with them. In response to these concerns and considerations, some countries have enacted legislation in this area. The method of protecting privacy has been proposed in a variety of ways, however most of them lack a precise mathematical and formal definition. In the end, there is no guarantee that this will ensure that the privacy of individuals will be protected in this way. In spite of the adoption of these methods, privacy violations have been reported from time to time. There was an introduction of Differential Privacy in 2006, which is a new concept that, through the use of a formal definition, ensures the preservation of a specific level of privacy and has gained considerable popularity among researchers and industry practitioners. In this definition, a central curator is assumed to be trustworthy, however, this is not always the case. Thus, a local differential privacy definition was proposed as a means of addressing users' distrust of central curators and to alleviate curators' concerns about privacy laws and the disclosure of data to third parties. There are several major companies who have adopted this marketing approach as a part of their privacy preserving strategies. These companies include Google, Microsoft, and Apple. A level of privacy for all users is considered by all methods of differential privacy mentioned above, including local differential privacy. Thus, some users may consider less privacy, resulting in the inevitable removal of their data, while others may experience more privacy. The purpose of this thesis is to reduce data noise and thereby increase the efficiency of local differential privacy by personalizing it. The solution can effectively address a wide range of user concerns. The evaluation of our method relies on metrics like mean absolute error, mean squared error, and other relevant measurements
- Keywords:
- Personalization ; Approximate Differential Privacy ; Pure Differential Privacy ; Local Differential Privacy ; Central Differential Privacy ; Database ; Data Analysis
Digital Object List
محتواي کتاب
Bookmark