Phoning It In

Using Telecom Data to Map Social Distance

Human society fractures along lines defined by politics, religion, ethnicity, and language. These cracks can lead to conflict and stifle efforts to advance social justice and economic development.

David Meyer, UC San Diego professor of mathematics and faculty affiliate with the University of California Institute on Global Conflict and Cooperation, leads an interdisciplinary team that develops new ways to understand social systems based on patterns they extract from large sets of data, such as cell phone records.

Grid Census Population
Cellphone calls map the population of Milan in near real time and at a finer scale than a traditional census. The ability to track shifts in density and distribution of people over the course of a day could help the city address problems with traffic and air quality, for example. Telecom data can be resolved to the size of a city block—the many blue squares shown on this map that fall within broader census tracts, outlined in black.

Phone traffic can reveal connections that other surveys miss and provide a measure of human density and distribution at a scale far finer than the size of a census tract, for example. Urban planners, social service agencies, human rights organizations, and health-care workers could use this type of information to identify neighborhoods where help is most needed.

In 2014, when Telecom Italia made available records for six hundred million calls placed in metropolitan Milan as part of a big-data competition, Meyer’s team used the information to paint a picture of immigrant populations in the city. “(Dis)assembling Milan,” their contest entry, made it into the top ten of 652 submissions.

Meyer’s team previously showed that mobile phone data could provide critical information about language communities and social divisions in regions where such divisions haven’t been well described. Their demonstration won first prize for best scientific project in the Data for Development Challenge, a competition held by the French telecommunications company Orange in 2013. They validated their approach in Côte d’Ivoire, where French is the national language, but more than seventy regional languages are also spoken.

Connections mapped among 1,216 cell phone towers for five months revealed “communities” of antennas—a kind of map depicting who talks with whom throughout the West African country.