Juiz online no ensino de CS1: requisitos, dificuldade de problemas e plágio em código-fonte

• Ano de defesa: 2016
• Autor(a) principal: Francisco, Rodrigo Elias
• Orientador(a): Ambrósio, Ana Paula Laboissière
• Banca de defesa: Ambrósio, Ana Paula Laboissière, Longo, Humberto José, Matos, Fernando Barbosa
• Tipo de documento: Dissertação
• Tipo de acesso: Acesso aberto
• dARK ID: ark:/38995/001300000b8c2
• Idioma: por
• Instituição de defesa: Universidade Federal de Goiás
• Programa de Pós-Graduação: Programa de Pós-graduação em Ciência da Computação (INF)
• Departamento: Instituto de Informática - INF (RG)
• País: Brasil
• Palavras-chave em Português: Juiz online; Ensino de programação; Plágio; Dificuldade de problemas
• Palavras-chave em Inglês: Online judge; Programming education; Plagiarism; Problem difficulty
• Área do conhecimento CNPq: CIENCIAS EXATAS E DA TERRA::CIENCIA DA COMPUTACAO
• Link de acesso: http://repositorio.bc.ufg.br/tede/handle/tede/6203

This dissertation approaches Online Judge in teaching Introductory Programming (CS1). Initially there was an exploratory research on BOCA system in teaching CS1, which brought experiences and data of student interactions with the system, which, supported by Systematic Literature Review (RSL), contributed to defining the requirements for the system meets the discipline of CS1 and guided continuing research. In the second phase, there was the aim to solve specific problems identified in the previous phase, and measuring the difficulty of CS1 problems and support for plagiarism identification in CS1 activities. The solution of these problems included RSL, practical experiences with writing and execution algorithms, comparison of the results with the expected results, and comparison of the proposed approaches to the identified in the literature. The strategy to measure the difficulty of problems CS1 proposed works with the height of a tree mounted to sets and sub-sets of nested code into a program and the amount of related subjects. The strategy to support the identification of plagiarism proposal works with the Edit Distance algorithm processing and normalization techniques in preprocessing, and it is a highly adapted proposal to the reality of the data used in this research (programs written in C with few lines of code by students CS1). Experience has shown the complexity of applying computing to education, which often works with subjective data, it was necessary to raise the difficulty of the problems in view of the students and the teacher’s view of the existence of plagiarism in peer programs, whose views are quite variables. It is suggested the creation of multidisciplinary teams to the evolution of the area (with professionals of computing, statistics, psychology and pedagogy) with a focus on validation and method used for research.