In order to bring about a world free of poverty, researchers and policy-makers would like to have a geographically-disaggregated poverty indicator system available to them. Such a system would enable a better understanding of the spatial distribution of poverty, and consequently result in a more efficient targeting of the poor. Typically, setting up such a system entails the collection of a heterogeneous geographically referenced poverty data since poverty is a complex social phenomenon, as well as the use of a Geographic Information System (GIS) for integrating the poverty data. Here, we discuss the use of the Self-Organizing Map (SOM) for analyzing various comparable sub-national poverty and welfare indicators across several time periods.