This paper aims to provide a tool to assist rural energy planning in developing countries. Indeed, classical methods seem to be limited for the implementation of power generation plants in these often remote and isolated areas. These localities are home to populations living on less than USD2 (2$) per day with a typical and complex conditions for the implementation of power plants. Thus, on the basis of surveys carried out in several villages in the northern region of Madagascar, a data classification method is used in order to provide electrification operators with a tool to help them make decisions on the potential technico-economic and cultural feasibility of electrification projects. The implicative and cohesive hierarchical classification method according to the interestingness measure (ICHC-), which has already proven itself in several didactic studies, is used herein for an economic-socio-energy study of households in some rural areas of the DIANA region of Antsiranana, in order to highlight the most used and consumed sources of energy in rural areas of the northern area of Madagascar. The results are conclusive because they highlight the potential of the sites in terms of electrification. Also, the approach makes it possible to make the link between the economic activities, the cultural practices of the villages, and the various elements relating to the energy component.