Introduction
Papua New Guinea (PNG) faces challenges in estimating the geospatial distribution of population poverty due to data availability and quality. The last census was conducted in 2011, and the next census has been delayed until 2024. Conducting field surveys in Papua New Guinea can be challenging due to various factors such as the country's rugged terrain limiting accessibility, low population density in many areas, the high cost of transportation and limited infrastructure.
Papua New Guinea is a culturally diverse country with over 800 languages spoken. Language barriers can make it difficult to gather accurate information. Additionally, security is an issue in some areas, where there is ongoing conflict or political instability. Thus, data collected by Mobile Network Operators (MNOs) represent an attractive non-traditional data source to estimate poverty in PNG.
In this paper, we investigate the feasibility of correlation between Digicel customer segmentation metrics based on mobile phone usage routinely collected for each subscriber for commercial purposes, and the poverty
status of subscribers as estimated by the World Bank high-frequency phone survey.
Read the full abstract
Preferred citation
Flowminder Foundation, Veres G., et al. (2023). Challenges in predicting individual poverty status from MNO customer segmentation metrics and phone surveys: a Papua New Guinea case study. https://doi.org/10.5281/zenodo.8414522
This abstract was accepted and presented at Netmob 2023.
NetMob is the primary conference on the analysis of mobile phone datasets in social, urban, societal and industrial problems.