A Data Mining Study for Distribution of Social Assistance Services: Sample of Yesilyurt Municipality

Abstract

Unemployment, poverty and similar problems that have come to the fore with the increase in population in our country have caused the municipalities to take charge in the field of social assistance and social services. For this purpose, it is very important that the municipalities that undertake social assistance and social service tasks are able to use the present data quickly during distribution of social assistance and make the right decisions in the related commissions. In this paper, data mining methods were used to establish a decision mechanism for the distribution of social assistance on the Yesilyurt Municipality, which is the district municipality of Malatya Metropolitan Municipality. In this study, Naive Bayes, Bootstrap Aggregating, Random Forest, Sequential Minimal Optimization and Locally Weighted Learning classificiation algorithms were used for the classificiation. As a result, Local Weighted Learning algorithm has the best result with 81.8%.