Allowing Finer Control Over Privacy Using Trust as a Benchmark

Abstract

Every time a user conducts an electronic transaction over the Internet a wealth of personal information is revealed, either voluntarily or involuntarily. This causes serious breach of privacy for the user, in particular, if the personally identifying information is misused by the other users present in the network. Ideally, therefore, the user would like to have a considerable degree of control over what personal information to reveal and to whom. Researchers have proposed models to allow a user to determine what personal information to reveal while doing a transaction over the Internet. However, these models do not help the user in determining who to trust, how much to trust and why to trust them with the personal information. The models fail to address loss of privacy through the misuse of information. In this paper we propose a privacy enhancing trust model to measure the degree of confidence that a user can have in the context of preservation of her privacy during a transaction. The model considers several factor while computing trust which include a user's own experience and knowledge about the target user and feedback obtained from groups of peer users called 'trusted neighbors' and 'friends'. The proposed scheme provides a flexible and powerful approach for the secure handling of private data and offers a user considerable control over how she wishes to disseminate her personal data