Um modelo para recomendação de cursos de especialização baseado no perfil profissional do candidato

• Autor(a) principal: Souza, Antonio Eduardo Rodrigues de
• Data de Publicação: 2013
• Tipo de documento: Dissertação
• Idioma: por
• Título da fonte: Biblioteca Digital de Teses e Dissertações do Mackenzie
• Texto Completo: http://dspace.mackenzie.br/handle/10899/24347
• topic: sistemas de recomendação; mineração de dados; aprendizagem de máquina; árvores de decisão; aconselhamento acadêmico; recommender systems; data mining; learning machine; decision trees; academic counseling; CNPQ::ENGENHARIAS::ENGENHARIA ELETRICA

Economic globalization has made products and services markets more competitive, demanding a better qualification of manpower. Consequently, companies are in need of best qualified professionals to meet specific demands. In this context, specialization courses options have been sought by professionals to acquire and update knowledge. However, diversification of courses offered by various institutions in various areas, targeted to specific audiences or general, or the lack of objective information, hinder the understanding of the important factors in the decision to be taken by a candidate. A poorly chosen option may incur factors that can trigger the change or even dropping out of the course. The topic is current and relevant to Higher Education Institutions (HEIs), showing the importance of offering specialized courses that are aligned to the skills of educational institutions and the interests of the professional training and retraining. Therefore, this paper proposes to study professional factors that influence candidates in choosing a course, and develop a recommendation model, using artificial intelligence techniques to practical use in HEIs, which assists applicants in the choice of courses, as well as serve as support and guidance to staff in the selection of candidates. It was applied a methodology based on processes Knowledge Discovery in Databases (KDD) and Cross-Industry Standard Process for Data Mining (CRISP-DM) in the evaluation of the historical data of candidates freshmen at a private university in the city of São Paulo, and proposed a recommendation model, which will identify the most suitable course of a candidate's profile, using a technique of data mining based on decision trees for the discovery of relevant knowledge from database. The completion of this project has allowed to propose courses that would be more suitable to professional profiles of the candidates, based on the professional and educational historical information which were considered more important for the candidate selection. It is expected, therefore, that the counseling service will become more accurate and more responsive to the selection of candidates, assisting to reduce the number of abandonments, dropouts or changes in specialization courses offered by the studied university.