Privacy

Assistance, Please! Smartphone Users Prefer Help with Privacy Settings

• By Dian Schaffhauser
• 07/11/16

Sorting out privacy settings for apps on a smartphone is no job for a mere mortal. A research project at Carnegie Mellon University found that most people preferred the settings offered up by an automated "personalized privacy assistant" compared to the ones they could figure out for themselves.

Reported in the paper, "Follow My Recommendations: A Personalized Privacy Assistant for Mobile App Permissions," the field study worked with 84 people using Android devices. Nearly 4 in 5 recommendations (79 percent) made by the assistant were adopted by users. After the initial set-up the same users received daily "nudges," encouraging them to continue reviewing and modifying their settings; but that follow-up encouragement only resulted in changes to five percent of the originally recommended settings, suggesting that people found them "useful and usable."

The assistant worked by engaging in a "dialog" with the user to understand overall comfort level in granting permissions to apps in certain categories. Through those answers, the app identified a user profile that was the best fit for a given person. The app privacy configuration recommendations were based on that profile. One profile "cluster," for example, described a person who denied location and contact access of social and finance apps. Another profile denied phone call log permission more often.

The project found that compared to users who were part of a control group the majority of recommendations in the assisted group were accepted, and participants kept most of the settings through the course of the study.

As the researchers concluded in their paper, "Our results suggest that personalized privacy assistants can indeed help users better manage their mobile app permission settings."

While the experiment focused on automated assistance for mobile interactions, they added, "We expect that personalized privacy assistant approaches [could] also be applied to support privacy decision making in other domains where privacy configuration or awareness is an issue."

"It's clear that people just can't cope with the complexities of privacy settings associated with the apps they have on their smartphones," said Norman Sadeh, professor of computer science, in an article on the experiment. "And it's not just smartphone apps. The growing number of sensors and other smart devices that make up the so-called internet of things will impact privacy and make it even more challenging for users to retain control over their data and how it is being used."

Sadeh has long been involved in researching privacy settings for social sites. Last year, he led a project to understand the impact of "mobile app privacy nudging."

The results of the latest study were presented at the Symposium on Usable Privacy and Security (SOUPS) in Denver, where it won the "SOUPS Privacy Award." The paper begins on page 36 of the proceedings PDF file.