No abstract

In this paper we investigate the question whether users' personalities are good predictors for privacy-related permissions they would grant to apps installed on their mobile devices. We report on results of a large online study (n=100) which reveals a significant correlation between the user's personality according to the big five personality scores, or the IUIPC questionnaire, and the app permission settings they have chosen. We used machine learning techniques to predict user privacy settings based on their personalities and consequently introduce a novel strategy that simplifies the process of granting permissions to apps.

Intelligent retail stores like Amazon Go collect and process a large amount of shoppers' personal data to offer their service. In this paper we present Retailio, privacy management software that allows the customer to select the private data that should be accessible by retail stores. A privacy wizard helps the user to set her privacy settings, by using either a small informal privacy questionnaire or privacy measures extracted out of the user's Facebook posts for a machine learning-based prediction of user-tailored privacy settings. We conducted an expert interview to determine the different types of data that could be recorded in intelligent retail stores, and performed a user study to find out whether their disclosures correlate with shoppers' personalities. Retailio was evaluated in a validation study, regarding accuracy of the privacy wizard and user experience of the software. Our results show that there is a strong correlation between the IUIPC questionnaire and the data disclosure choice, which allowed us to predict the privacy settings with 70% accuracy.

This paper proposes to recommend privacy settings to users of social networks (SNs) depending on the topic of the post. Based on the answers to a specifically designed questionnaire, machine learning is utilized to inform a user privacy model. The model then provides, for each post, an individual recommendation to which groups of other SN users the post in question should be disclosed. We conducted a pre-study to find out which friend groups typically exist and which topics are discussed. We explain the concept of the machine learning approach, and demonstrate in a validation study that the generated privacy recommendations are precise and perceived as highly plausible by SN users.

No abstract