Resumen de Minería de datos para perfilamiento de las brechas digital y educativa de ciudades en censos de población y vivienda
Sergio R. Coria, Ivania Orozco, Juan Luna
This work proposes and evaluates a data mining methodology to discover profiles of cities on the basis of educational characteristics of inhabitants and on discrete classes describing the presence of information and communication technologies (ICT) in households. City profiling involves discovering the variables and their corresponding values that allow distinguish classes of cities within a country regarding a determined target variable. Pattern discovery on the interaction among educational attainment of inhabitants, presence of ICT in households and other demographical and economical variables is relevant to researchers, public policy makers and private company managers. Using city (municipality) as analysis and modeling unit is novel in this research field because most of previous work has used the country level. In addition, the data mining approach is also novel in this area because prevailing approaches are based on creating composite quantitative indexes or on performing multivariate analyses.
