Submitted by Automa??o e Estat?stica (sst@bczm.ufrn.br) on 2016-08-16T20:15:31Z No. of bitstreams: 1 LeandroDeAlmeidaMelo_DISSERT.pdf: 2505718 bytes, checksum: c3f7b02d7636033873d55d92a37d5e28 (MD5) / Approved for entry into archive by Arlan Eloi Leite Silva (eloihistoriador@yahoo.com.br) on 2016-08-17T22:38:52Z (GMT) No. of bitstreams: 1 LeandroDeAlmeidaMelo_DISSERT.pdf: 2505718 bytes, checksum: c3f7b02d7636033873d55d92a37d5e28 (MD5) / Made available in DSpace on 2016-08-17T22:38:52Z (GMT). No. of bitstreams: 1 LeandroDeAlmeidaMelo_DISSERT.pdf: 2505718 bytes, checksum: c3f7b02d7636033873d55d92a37d5e28 (MD5) Previous issue date: 2016-02-22 / Coordena??o de Aperfei?oamento de Pessoal de N?vel Superior (CAPES) / Habilidades transversais e pr?ticas de desenvolvimento em projetos foram identificadas como as principais defici?ncias dos egressos de cursos de computa??o. Essa problem?tica motivou a realiza??o de uma pesquisa qualitativa sobre os desafios encontrados por professores desses cursos na condu??o, acompanhamento e avalia??o de projetos colaborativos de desenvolvimento de software. Dentre os desafios identificados, destacam-se as dificuldades para acompanhar e avaliar a participa??o dos alunos em projetos acad?micos. Nesse contexto, uma segunda pesquisa de natureza quantitativa foi realizada com o objetivo mapear habilidades transversais dos alunos a um conjunto de indicadores que podem ser extra?dos a partir de reposit?rios de software usando t?cnicas de minera??o de dados. Tais indicadores visam auxiliar o professor no acompanhamento de habilidades transversais, tais como: a participa??o no trabalho em equipe, a lideran?a, resolu??o de problemas e o ritmo de comunica??o durante projetos. Para isto, uma abordagem de avalia??o por pares foi aplicada em uma turma de desenvolvimento colaborativo do curso de Engenharia de Software da Universidade Federal do Rio Grande do Norte (UFRN). Essa pesquisa apresenta um estudo de correla??o entre os scores das habilidades transversais dos alunos e os indicadores baseados na minera??o de reposit?rios de software. O objetivo da pesquisa ? melhorar a compreens?o das din?micas de trabalho em projetos colaborativos de estudantes, assim como incentivar o desenvolvimento de habilidades transversais que s?o exigidas pela ind?stria de desenvolvimento de software. / Soft skills and teamwork practices were identi ed as the main de ciencies of recent graduates in computer courses. This issue led to a realization of a qualitative research aimed at investigating the challenges faced by professors of those courses in conducting, monitoring and assessing collaborative software development projects. Di erent challenges were reported by teachers, including di culties in the assessment of students both in the collective and individual levels. In this context, a quantitative research was conducted with the aim to map soft skill of students to a set of indicators that can be extracted from software repositories using data mining techniques. These indicators are aimed at measuring soft skills, such as teamwork, leadership, problem solving and the pace of communication. Then, a peer assessment approach was applied in a collaborative software development course of the software engineering major at the Federal University of Rio Grande do Norte (UFRN). This research presents a correlation study between the students' soft skills scores and indicators based on mining software repositories. This study contributes: (i) in the presentation of professors' perception of the di culties and opportunities for improving management and monitoring practices in collaborative software development projects; (ii) in investigating relationships between soft skills and activities performed by students using software repositories; (iii) in encouraging the development of soft skills and the use of software repositories among software engineering students; (iv) in contributing to the state of the art of three important areas of software engineering, namely software engineering education, educational data mining and human aspects of software engineering.