296 / 2017-02-04 17:50:42

Fine Grained Privacy Measuring of User’s Profile over Online Social Network

12752,12753,12754,6136,12755

In recent years, online social networking has become part of every ones life. A person known to computer has its online social accounts like Face book, twitter or MySpace. A large body of work has been devoted to address privacy concerns related to social networks. Many authors have discussed about to share social networks without revealing the identities or the sensitive relationships of the users involved. But a lot attention is needed to identify to the privacy risk of users posed by their daily information sharing activities. There is no unified theory of privacy and its measurement. This paper focuses on obtaining a practical approach to quantify measure and evaluate privacy. This paper, presents privacy issues raised in online social networks from the individual user’s viewpoint than propose a framework to compute the privacy index for a user on OSN and show the applicability and requirement of privacy index. This score indicates that the user is aware of their privacy profile or not. The index can be used for the recommendation to enhance the privacy settings of the users in the group. Framework proposes a mathematical model of basic commodity index to calculate privacy index of any user in OSN. Our definition of index satisfies that more sensitive information a user discloses, the higher his or her privacy risk so its index number. The framework considers both sensitivity and visibility of information of user’s profile and computes index value on the basis of them. The sensitivity of profile items over survey data is calculated using naïve formula. Based on privacy measurement function values the users on OSN are classified in three categories as Secure, Mediocre and Vulnerable to privacy attack. We further compare the normal indexing technique to our privacy measurement function. This along with propose algorithm shows better efficacy and reduces the possibility false classification of user's categories.