Paper Reading #17: Privacy Risks Emerging from the Adoption of Innocuous Wearable Sensors in the Mobile Environment

Andrew Raij is an associate professor of electrical engineering at University of South Florida
Animikh Ghosh is a junior research associate at SETLabs.
Santosh Kumar is an associate professor at the University of Memphis
Mani Srivastava is an associate professor of electrical engineering at UCLA.

This paper was presented at CHI 2011.

Summary


Hypothesis
Researchers set out to prove, in this paper, that a users personal privacy can be violated by inferring data from mobile sensors.

Methods
Researchers developed a framework that allowed them to gain information from sensors and infer other data. By receiving raw data from a sensor, the information can be reconstructed to decipher behaviors and discover contexts.
An example they gave related to a woman who was jogging and has a seizure. Through sensor movement data, they could infer she was running and then that she was having seizure. They could also discover information about the context well (such as where she was jogging as well as time).

Due to this information being able to be inferred, researchers believed that this could lead to privacy threats (financial, psychological, and physical).

To confirm their thoughts, researchers set out to test three goals: discover how much people care about privacy concerns, use the framework to analyze behaviors, discover how identifications of the data effects concert.

To conduct this experiment, they used two groups. The first group had "no personal stake". They did not wear an AutoSense sensor. They were simply asked a questionnaire. The second group had a sensor and were also asked the same questionnaire.
The sensor group was asked to wear the AutoSense sensor during their awake hours for 3 days.

Results
At the end of the experiment, they were able to use data collected by AutoSense to analyze the sensor group's three days.

In the questionnaire, they were able to discover how concerned people were about given factors. The sensor group was much more concerned following the test. People said that they'd be concerned about people knowing their location at a given time.

The participants also expressed concern about who the results were shared with. They expressed much more concern if their identity was shared with the data.

Discussion

While parts of this paper were generally hard to follow, it was still an interesting concept. Many of our mobile devices are capable of sensing movement and audio. If our mobile devices started collecting information through the day about our movements, certain things might be extrapolated and thus might raise some privacy concerns.

One of the large problems in the coming age of technology is making sure this kind of sensor data does not get compromised or leak out. Especially if systems that analyze this data are created. This could become an entirely new branch of cybercrime.