Using Big Data to Predict Poverty Index

By Chris Orwa
Data Science Lab
  Published 06 May 2016
Share this Article

The iHub Research Data Lab will be holding its May Data Science Meetup on the 13th of May at the iHub UX Lab from 5.30pm. This month’s meetup will be presented by Christopher Wambugu Njuguna, an aspiring data scientist and researcher. He received a B.S. degree in Computer Science from Africa Nazarene University and an M.S. in Information Technology from Carnegie Mellon University in Rwanda. He has worked for over ten years in the information technology field in Kenya and abroad. His fledgling research career includes spatial-temporal poverty level estimation using big data and inferring spatial-temporal human mobility from call records.

Big data offers the potential of calculating timely estimates of the socioeconomic development of a region. Mobile telephone activity provides an enormous wealth of information that can be utilized along with traditional household surveys. Estimates of poverty and wealth rely on the calculation of features from call detail records (CDRs). However, mobile network operators are reluctant to provide access to CDRs due to commercial sensitivity and privacy concerns.

Christopher will be presenting on using big data to calculate a multi-dimensional poverty index.As a compromise, we show that a relatively sparse CDR dataset combined with other publicly available datasets based on satel-lite imagery can yield competitive results. In particular, we build a model using two features from the CDRs, mobile ownership per capita and call volume per phone, combined with normalized satel-lite nightlight data and population density, to estimate the multi-dimensional poverty index (MPI) at the sector level in Rwanda. Our model accurately predicts the MPI for sectors in Rwanda that contain mobile phone cell towers.

Register for the event here:


comments powered by Disqus