Atnafu, Solomon (PhD)Feyiso, Firesew2019-09-262023-11-292019-09-262023-11-292018-06-05http://etd.aau.edu.et/handle/123456789/19203An increase in the number of residence with an increase in urbanization results a growing demand of quality services. The geographical relationship of infrastructures,economic and political aspect of urban region is requred to manage and plan the activities of service facilities. Lack of all-inclusive site selection with the consideration of geographical relationship of the existing infrastructure, population distribution and density, natural and artificial barriers, legal and political aspect of the location result improper distribution and allocation of resources for the public and miss use of the existing services facilities. Site selection in Addis Ababa usually depends on the existence of free and cheap land. The purpose of this study is to design and develop a model that can identify and support decision makers to select optimal service facility site using suitability analysis and data mining outputs for Addis Ababa. Our model employed unsupervised classification of data mining called spatial clustering and suitability analysis using GIS-based multi Criteria Analysis for optimal site selection. First, suitability of the study area will be identified using carefully selected and weighted factor and constraint criteria by pair wise comparison using analytical hierarchical process, which is the systematic engineering method to quantitatively analyze non-quantitative objects. Then spatial data mining using road data identifies clusters and cluster center. Finally suitability map and data mining results are used to select candidate location and optimal site selection using different kind of vector and image maps. The model implemented using C# and is tested for hospital site selection using Nifasilik Lafto sub city, Addis Ababa and suitability map of Addis Ababa for hospital site selection. The overall optimal site selection accuracy is tested and evaluated by experts and it has been found that different candidate site and optimal site is selected. The overall performance of the proposed model from summary of validation report is 85.2%. And the system can provide adequate justification or reason for the selected site is about 88% effective.enSpatial Data MiningSuitability AnalysisOptimal Site SelectionCandidate LocationGIS-Based Multi Criteria AnalysisAddis AbabaThesis