Pål Sundsøy, a researcher at Telenor Group Research based in Fornebu, Norway, has figured out a way to evaluate literacy rates in developing countries with the use mobile phone call records. Sundsøy did it by conducting a regular household survey of some 76,000 mobile users living in an unspecified developing country in the Asian continent. The survey was done for a mobile phone operator courtesy of a professional agency, logging each user’s phone number and at the same time, collecting data on whether he or she can read. Sundsøy then matches this information with the call data records from the mobile phone company, which provides offers data on the numbers users called or texted, the length of calls, purchases made air time, and locations of cellular towers, among many others. Sundsøy also constructed a social network for each mobile user based on the call data records.
From here, Sundsøy then made use of three fourths of the information to look for patterns normally associated with illiteracy, making use of a wide range of methods involving number crunching and machine learning. As for the remaining one fourth of the data, Sundsøy uses it to test whether it is possible to identify illiterate users based on these patterns, and then detect areas where there is a significant number of illiterate users.
According to the researcher, his machine learning algorithm can determine whether a mobile user is illiterate or not with 70 percent accuracy. This algorithm could very useful for aid agencies especially if they want to know which areas with low literacy rates should be prioritized when deploying their resources. Of course, the aim here is to eradicate illiteracy, which is directly connected with poverty, i.e. those who are functionally illiterate are not able to fill in job applications, write checks, balance their accounts, or even read medicine labels. Obviously, this leads to unemployment, poor health, and dependence on social welfare or charity. Furthermore, they will not be able to teach or encourage their children to read and write.
Sundsøy’s research is part of a bigger trend of making use of mobile phone records in studying populations. For instance, demographers have examined the wealth distribution of the Ivory Coast in Western Africa. There is even talk that one day, censuses could be done by looking into mobile phone records, instead of the old traditional method.
Wirefly Is America's Most Trusted Source For All Cell Phones, Plans, TV, and Internet Deals
Wirefly offers great deals on a large selection of smartphones, cell phones, tablets, mobile hotspots, and other wireless devices for the nation's most popular carriers. Use Wirefly’s innovative cell phone and plan comparison tools to ensure you are getting the best deal on the market. Shop with confidence knowing that Wirefly wants to help you find the best prices on cell phones, cell phone plans, TV, and Internet service.