Cell-Phone Technology and Data Analysis Featured Pattern: P0845 November 2015

Author: Barry Jones

Cell-phone technology and infrastructure have unexpected uses in data analysis.

Abstracts in this Pattern:

Falling rain creates signal interference among cell-phone towers, and researchers from the University of Ouagadougou (Ouagadougou, Burkina Faso) found that this interference can provide a rainfall measure that is much more accurate than those of radar, satellites, and conventional pluviometers. Rainfall-measuring methods typically in use suffer from accuracy and availability limitations—for example, satellite systems lose accuracy at certain times and in some locations, and many countries cannot afford radar installations. In contrast, cell-phone towers are ubiquitous even in many developing countries, and the cell-phone-signal interference provides a highly accurate measure of rainfall.

Researchers at the Massachusetts Institute of Technology (Cambridge, Massachusetts) discovered that cell-phone data can indicate changes in a cell-phone user's employment status. The researchers conducted a study and identified how people's communication patterns change after they have stopped working. For example, after a person has been laid off, he or she receives fewer incoming phone calls, contacts fewer people each month, and contacts a different set of people.

In developing countries, the financial information typically in use to assess the likelihood of a loan applicant's defaulting on a loan is often unavailable. Recently, scientists at Brown University (Providence, Rhode Island) discovered that cell-phone usage can be indicative of creditworthiness. During their study, the researchers analyzed cell-phone data from the 90 days leading up to a person's receiving a loan. The researchers identified several behaviors—for example, maintaining plenty of prepaid minutes and calling a large number of people—that correspond with the likelihood a person will repay a loan.

The accuracy of inferences based on a technology's secondary uses is always at risk because of technological changes and incorrect assumptions; however, the growing amount of data that devices record will enable even more analytical possibilities—especially as the Internet of Things connects an increasing number of devices and therefore data sources.