K-anonymity against Neighborhood Attacks in Weighted Social Network

• Main Authors: Wun-ShengYao, 姚文盛
• Other Authors: Jung-Shian Li
• Format: Others
• Language: en_US
• Published: 2014
• Online Access: http://ndltd.ncl.edu.tw/handle/6qa7ju

碩士 === 國立成功大學 === 電腦與通信工程研究所 === 102 === With the Internet, mobile and the information expanding rapidly, more and more social network data is provided for research in the SoLoMo times. So the personal privacy protection in social network is an important issue. SoLoMo means to integrate the elements of social, local and mobile effectively, it combines the virtual network and the real world. The problem incident to the times is that the relations between you and your social groups are easy to be revealed, and it gets worst in the weighted social network. To get the optimal decision from the analysis of the big data, the key is the truth of information. Especially, it is a big issue to get the balance between privacy protection and data usability while the information involving some personal privacy. For example, we usually add virtual relations to achieve K-anonymity protection in order to solve neighborhood attacks. The thesis is different than others before it. Our goal is to add virtual edges as less as possible, and furthermore, we also changed less weights to achieve k-anonymity. Firstly, processing the more important and easily revealed information, and we handle the similar data at the same time in order to reduce adding virtual relation and increase the data usability.