Skip to main content

Detecting Networks using Mobile Phone Data

In this era of smart phones, consumers are connecting with each other using their smartphones. Although on a micro level one phone call being made may not have much to say about a network but on a macro scale, the phone data provides indicators about the connectivity of people across cities and localities. In recent years data from mobile phones has provided indicators for commuting patterns in major cities, wealth distribution in African countries and reproductive strategies in Western societies. This provides insightful data for economists, sociologists and networks. Daniel Kondor and companies at SENSEable City Laboratory, part of MIT, and at Ericsson, have developed a mathematical tool called ManyCities that presents mobile data in three simple ways. The first shows how phone usage over time, revealing clear daily and weekly patterns as well as longer term needs. It shows how activity clusters in different parts of cities and provides analysis of residential and business areas.

This links to our networks class as we have studied the sizes of network components and their linkages. For instance, mobile phone sets could be nodes and text messages could be edges present between them. When we thought about isolated networks such as North Korea a question to consider whether there are even text messages coming in and out of the country. In such a network of mobile phone data, a question to consider is whether any network can ever be isolated? Another possible application could be whether links between clusters could be indicators of communications across different major centers ranging from the links between New York and London to links between two neighboring colleges. The data obtained from ManyCities technology can potentially revolutionize the study of networks as it can provide information on what has become an intrinsic part of human behavior- access to the internet as well as cellular communication between people.




Leave a Reply

Blogging Calendar

September 2015