Location-based Services: Raising the Bar on Privacy

March 17, 2009

You just finished a productive day at a conference, and you're ready to meet up with some potential clients. With ease, your Smartphone displays a map showing a pub with a four-star happy hour within two blocks. Just as you're falling in love with your gadget all over again, a tweet from a nosey former classmate says that he'll meet you there! No doubt, he will offset the tone of your evening, and he might even promote his own company to your new contacts.

While your mobile device can be the cure when you urgently need information, some people complain about an unwanted side effect: a lack of privacy.

Geo-tracking, just described, is an example of mobile location-based services (LBS) that has raised privacy issues. LBS send a user's current location to other users or service providers in order to find restaurants, gas stations and other essentials in their vicinity, but it also ends up telling them exactly where the user can be found--not the best tool for a private person. Working with a team at Carnegie Mellon, INI student Rahul Raheja is designing a better application that would appeal to the privacy-conscious person. 

His project addresses two major concerns with LBS. First, LBS require that the user be connected online, but a connection is not always available, and when it is, it drains battery life. Second, by mapping your location online, you can all too easily have your privacy invaded, if not by an outright stalker, then by others whose curiosity is just generally unwanted.

Rahul explains that about a third of the mobile population is greatly concerned with protecting their privacy online, while another third is not concerned at all. His aim is to provide a usable solution that serves the need for all kinds. Using his team's application, a user can specify the granularity of his location he is willing to reveal by adjusting the privacy settings in his profile, such as his city for a high level of privacy, or his zip code or street address for a low level.

"No application meets the privacy needs of all kinds of users adequately. But we want to provide an application that does consider the importance of privacy as a factor," said Rahul, who is a second-year student in the Master of Science in Information Networking program.

The project investigates the use of pre-fetching data, where the device loads and saves location-based information ahead of time. When the user needs it later on the go, even if an Internet connection is unavailable, the cell phone will retrieve and map nearby "points of interest," such as dining, theaters, book stores, and other useful places as determined by the service provider. Through the business model the team devised, the user would be able to select the type of service provider they wish, according to their interests. If the user were to choose Zagat as a service provider, for example, the application would deliver nearby restaurants and their ratings.

The first inspiration for this project came from a faculty member, Professor Jason Hong, who is Assistant Professor with the Human-Computer Interaction Institute. He mentored Rahul and three other students in the course Mobile and Pervasive Computing. Three students stayed on the project after the semester-long course ended. They are currently working on a conference paper to be submitted at UBICOMP 2009, a ubiquitous computing conference to be held in Orlando, Florida later this year.

"This project will give a head start to deploy usable LBS applications that respect one's privacy," said Rahul hopefully. "We will see where it goes from here."