Authentication based upon secret knowledge and its resilience to impostors

This dissertation presents an assessment made on an alternative to the present password and PIN-based methods of user authentication. In the recent years, many alternative authentication methods emerged, but none of them seems to have been a major breakthrough. Nevertheless, two techniques emerged as potentially efficient: image-based authentication and cognitive questions techniques. Even if the viability of these techniques has been proved, rare are the researches that assessed the resilience of the methods to impostors. Therefore, an environment has been created to test the robustness of the alternative techniques, and rendered available to the widest possible audience. The evaluation comprises both a theoretical and a pragmatic analysis to rate the robustness of the methods, and produced relevant statistics and results. In order to analyse the data collected during the experimentation, a dedicated program has been created. The results show that the methods are differently vulnerable to the practical tests, with an outstanding success rate for the phishing test. In the other side, Social Engineering was inefficient for PassImages when it works amazingly well for cognitive questions