AN XML-BASED GENETIC ALGORITHM TOOL – SYSTEM FEATURES AND MODELING Amílcar Soares Júnior Gustavo Henrique Matos Bezerra Motta Celso Augusto Guimarães Santos Cristiano das Neves Almeida amilcarsj@gmail.com gustavo@di.ufpb.br celso@ct.ufpb.br almeida74br@yahoo.com.br Center for Technology, Federal University of Paraíba 58051-900 João Pessoa – PB – Brazil Camilo Alysson Simões de Farias Koichi Suzuki camiloallyson@yahoo.com.br ksuzuki@dpc.ehime-u.ac.jp Department of Civil and Environmental Engineering, Ehime University 790-8577 Matsuyama, Ehime, Japan Abstract. In many different fields of science we are faced with optimization problems. Several approaches have been proposed to solve such difficulties, including the use of Genetic Algorithms (GA). This paper aims to present the features and the modeling of a configurable and portable GA tool that uses the eXtensible Markup Language (XML) to describe input and output files for an optimization problem. Techniques such as the Object-Oriented Programming Paradigm (OOPP) and Design Patterns are used in order to build the GA tool. The tool is applied for the calibration of the rainfall-runoff models in the Brazilian National Water Resources Information System (SNIRH, in Portuguese). Rainfall-runoff models are hydrologic mathematical models used to simulate the hydrologic cycle and, in the present case, a set of fictitious reservoir and parameters are used. The GA was used in this study to find the best set of parameters for each model. The calibration and test results show that the XML-based GA tool was very efficient for defining the rainfall-runoff model parameters. Since any application or platform capable of processing XML could utilize this tool, it may be considered as an efficient and effective alternative for solving optimization problems. Keywords: Genetic Algorithms, XML, Object-Oriented Programming Paradigm, Design Patterns.