Sign In

Communications of the ACM

ACM TechNews

How to Mine Cellphone Data Without Invading Your Privacy

An image of cellular call density in New York's Central Park.

One use of cellphone based mobility models is to understanding inter-city or regional movements, such as commuting or the patterns of movements to and from landmarks. Call density on a summer Saturday in New Yorks Central Park is shown here, with heavies

Credit: AT&T Research

Researchers at AT&T, and Rutgers, Princeton, and Loyola universities have found a way to eliminate personally identifiable information in cellphone data, which could significantly improve data-mining potential by removing privacy concerns.

The team used billions of AT&T phone call and text-message location data points to make a mobility model of Los Angeles and New York City. The model aggregates the data, generates synthetic call records, and mathematically conceals personally identifiable data.

The model can quickly forecast the impact of new developments or telecommuting policies on transportation, says Princeton's Margaret Martonosi. In addition, the model could aid in town-level planning in which minimal mobility data exists, with planners now limited to road sensors and the handful of people who allow the use of their GPS position data.

Call detail records (CDRs) are created by all phones and kept by mobile carriers, providing approximate user locations that can create an accurate trace of user movements over time. By offering an unprecedented picture of population movement patterns, aggregated CDRs could advance epidemiology research, ease traffic congestion, and guide development in developing countries. Eliminating privacy concerns removes the most significant barrier to using CDRs for research.

From Technology Review
View Full Article


Abstracts Copyright © 2013 Information Inc., Bethesda, Maryland, USA


No entries found

Sign In for Full Access
» Forgot Password? » Create an ACM Web Account