User Profiling: A Privacy Issue in Online Public Network

Abstract

The internet users, especially those who are using the Online Public Network (OPN) have concern about their private data. The OPNs are collecting user information from their profiles which are provided during the account creation and linking these attributes (for example age, gender and location) to the user explicitly posted data. In this way the OPNs collect and analyze the user’s psychological and behavioral characteristics to create user profiles. These user profiles are used for advertisement purposes and can be provided to third party for financial profit, but these profiles have private information about users which could be used for malicious purposes. In this paper, a relationship hiding technique is proposed to protect the user from profiling. The relationship between the user account profile data and the user explicit posted data has been hidden using the data encryption to protect user privacy, but the encryption hides the information even from the legitimate users. In order to allow the legitimate users to see the shared contents, a collaborative sharing technique is proposed. The collaborative sharing technique is implemented and tested as a proof of concept which gives satisfactory results. The proposed and the state of the art sharing techniques are taking almost same processing time.