An automated resource allocation system for software projects based on employee skills and project attributes
Students Name: Voitseshchuk Mariana Volodymyrivna
Qualification Level: magister
Speciality: Software Engineering
Institute: Institute of Computer Science and Information Technologies
Mode of Study: full
Academic Year: 2022-2023 н.р.
Language of Defence: англійська
Abstract: The purpose of the qualification work is to study the existing methods and models of automated distribution of human resources, optimization of the process of selection and distribution of IT specialists to the project, as well as the creation of a software system that helps IT project managers to make the right decision regarding the selection of IT specialists to the project. The work consists of four sections and three appendices: analysis of the task of automated selection of specialists for the project, analysis of solutions to the task of automated selection of specialists for the project, analysis of requirements for the system of automated selection of specialists for the project and its software implementation and practical results of the study of automated selection of specialists for the project . In the process of analyzing literary sources, the peculiarities of the task of team formation were investigated and the search-based optimization method was singled out as one of the most successful approaches to its solution. A comparative analysis of common search-based algorithms is performed, and a genetic algorithm (GA) is chosen because of its accuracy and efficiency in finding global optima. The software requirements were analyzed, the purpose of the system was worked out in detail, and its quality attributes were determined. The architecture of the system was developed, the logic of the program and the genetic algorithm was described, and their software implementation was carried out. Experiments were conducted to check the speed and efficiency of the algorithm. A survey of project managers and coordinators was conducted in order to assess the possibility of using the product for real work tasks. The possibilities of improving the product and its algorithmic component have been analyzed. In order to increase the accuracy of the algorithm and narrow down the search space, it is proposed to apply clustering methods. The work excluding appendices includes 57 pages