Professor Alfred Kobsa and the Research on Household Privacy Decisions

by Sharmin Shanur

Photo courtesy of Sharmin Shanur

In a petite room on the fifth floor of Donald Bren Hall, information and computer sciences professor Alfred Kobsa and five PhD students come together weekly to discuss ways in which they can merge the information-hungry world of technology with the burgeoning need to maintain individual privacy. To put it simply, they are trying to create personalized privacy settings in all electronic devices—in particular, home security devices.

Each of the five PhD students under the mentorship of Professor Kobsa are working on a different aspect of privacy in the tech world: Eugenia Rho, an Informatics major, is studying the effects on privacy of self-driving cars and Yao Li, also a third year Informatics major, is conducting cross-cultural privacy research in which she explores how international companies must change their privacy expectations from country to country.

“Individualized countries are okay with having data collection from computers, but do not feel the same for their phones. However, collective societies are okay with data collection through their phones but not through their computers,” noted Li.

Companies that do business in both types of environments must understand the extent to which they are allowed to collect information from their patrons.

For Rho, her focus is to specifically examine the technology used inside self-driving cars. What draws the line between intrusion of privacy and convenience?

To make her claims more understandable, Rho gave an example of how Nissan has a relationship with Verizon, a third-party provider which dictates the GPS and calling systems, along with other things, in cars built by Nissan. This allows Verizon to collect individualized information about each driver, such as where their work is located, when they go to the gym, where they are at night, and more. Of course, the GPS system is built for convenience, but where can a person decide when a system is convenient or intrusive? Well, Rho explores this question in her paper titled Privacy Norms in the Context of Connected and Self-Driving Cars.

In this paper, she “highlight[s] key privacy challenges and issues in the context of connected and self-driving cars.”

Rho is certainly excited about how cars are becoming “smarter,” but remains fearful about the implications it may have on our individual privacy.

As much as each individual project conducted by the students is interesting and helps put UCI at the forefront of tech research, the most exciting aspect for Informatics students at UCI is the  large grant from NSF (National Science Foundation) Professor Kobsa received to do research on household privacy decisions.

Having recently received this grant, Professor Kobsa and his team are at the forefront of designing the plan and the research has yet to begin. However, Professor Kobsa indicated that if one measures the energy consumption of an individual person repeatedly, one can know what that the energy consumer is doing on a daily basis.

He humorously explained that through the Edison electricity system present in his house and many other houses, he can tell you “what his wife is doing, [and] who she was with last night.” This knowledge is certainly powerful because every single time  his wife turns on  the T.V. or stove, she creates a permanent mark on the energy system. However, Professor Kobsa believes that a person should be able to decide what sort of information goes to an electricity provider.

Therefore, he and Hosub Lee, one of the five PhD students majoring in informatics, published a research paper about a system that they created with a 77 percent accuracy in predicting what sort of privacy setting an individual desires. This research took place throughout UCI’s campus and included about 170 participants. By conducting a series of tasks through the aid of Google glasses to document each participant’s day-to-day life, Professor Kobsa and his team were able to place each participant in a generic cluster and predict their level of privacy.

These are the many experiments Professor Kobsa and his team are doing to introduce a new world of global technology. Although research like this takes time and money, Professor Kobsa is hopeful that they can begin to understand privacy on a global scale and hopefully create systems that maintain a person’s ability to dictate the sort of information that is dispersed about them.