A group of INESC TEC researchers was acknowledged due to their research work on the management of permissions on mobile devices. The team developed a set of techniques to automate the response to requests for permissions by smartphone applications, with a reliability of 90%. This work received the award for best scientific paper at the ACM CODASPY conference, which took place in the United States of America.
Currently, mobile devices allow users to control permission requests and access to the devices’ data and sensors. Users can choose whether to authorise said requests. However, many believe that current managers are not effective in protecting and notifying users to the privacy risks of their decisions. In this sense, a team of researchers from INESC TEC presented a solution that enables the safe prediction of privacy preferences; the results of this research work were published in the scientific paper Prediction of Mobile App Privacy Preferences with User Profiles via Federated Learning.
“Different methods for automating privacy protection on mobile devices have already been developed; however, they depend on the access to all of the users’ data. In this paper, we propose new methods that allow to predict privacy preferences in a distributed/federated way, without the need for a centralised access to the users’ data”, explained researcher João Vilela.
The recommended prediction methods make it possible to automate the smartphones apps’ response to requests for permissions, with a reliability of 90%. “The techniques developed are crucial to avoid manually checking close to 35 requests for permissions per hour, which would be the optimal solution in terms of privacy, but unfeasible in terms of usability,” he stated.
Regarding the paper’s acknowledgment, João Vilela perceives it as “the scientific community’s recognition of the need to increase privacy protection on mobile devices, while acknowledging the techniques for privacy protection automation that we have developed”.
The work reported in the paper was developed by André Brandão, a student of the master’s degree in Data Science at FCUP, under the guidance of Ricardo Mendes (University of Coimbra) and João Vilela.
The INESC TEC researchers mentioned in this news piece are associated with INESC TEC, UC-CISUC and UP-FCUP.
News, current topics, curiosities and so much more about INESC TEC and its community!
