Why That Picture? Discovering Password Properties in Recognition-Based Graphical Authentication

Numerous graphical authentication ideas have been proposed on how to address the security and usability of text-based passwords. However, it remains unclear how users approach graphical password selection and the inherent personal bias when selecting images. This study investigates user choices in password selection for recognition-based graphical authentication. Our analysis is based on a total of 302 participants continuously using a graphical authentication system during a 6-week long study. The results show pronounced preference effects for image properties such as color, shape, and category. Additionally, there is a significant difference between genders in the selected images based on the same properties.