The MSR Cookbook: Mining a decade of research

Hemmati, Hadi; Nadi, Sarah; Baysal, Olga; Kononenko, Oleksii; Wang, Wei; Holmes, Reid; Godfrey, Michael W.

doi:10.1109/msr.2013.6624048

Cited by 45 publications

(32 citation statements)

References 62 publications

Supporting

Mentioning

Contrasting

Unclassified

Order By: Relevance

“…Such traceability is complicated by the fact that the sources of bug data might not be always in sync or complete [5]. Development logs should form a superset of all development data: one would expect the data contained in issue trackers to be mirrored in the version control commit logs, and developers to record and distinguish their actions between development and bug fixing.…”

Section: Introductionmentioning

confidence: 99%

Towards an automation of the traceability of bugs from development logs

Romo

Capiluppi

2015

Proceedings of the 19th International Conference on Evaluation and Assessment in Software Engineering

View full text Add to dashboard Cite

Context: Information and tracking of defects can be severely incomplete in almost every Open Source project, resulting in a reduced traceability of defects into the development logs (i.e., version control commit logs). In particular, defect data often appears not in sync when considering what developers logged as their actions. Synchronizing or completing the missing data of the bug repositories, with the logs detailing the actions of developers, would benefit various branches of empirical software engineering research: prediction of software faults, software reliability, traceability, software quality, effort and cost estimation, bug prediction and bug fixing.Objective: To design a framework that automates the process of synchronizing and filling the gaps of the development logs and bug issue data for open source software projects.Method: We instantiate the framework with a sample of OSS projects from GitHub, and by parsing, linking and filling the gaps found in their bug issue data, and development logs. UML diagrams show the relevant modules that will be used to merge, link and connect the bug issue data with the development data.Results: Analysing a sample of over 300 OSS projects we observed that around 1/2 of bug-related data is present in either development logs or issue tracker logs: the rest of the data is missing from one or the other source. We designed an automated approach that fills the gaps of either source by making use of the available data, and we successfully mapped all the missing data of the analysed projects, when using one heuristics of annotating bugs. Other heuristics need to be investigated and implemented.

show abstract

Section: Introductionmentioning

confidence: 99%

Towards an automation of the traceability of bugs from development logs

Romo

Capiluppi

2015

Proceedings of the 19th International Conference on Evaluation and Assessment in Software Engineering

View full text Add to dashboard Cite

show abstract

“…As etapas típicas de um processo de MSR são [Hemmati et al 2013]: I) Extração, II) Modelagem de Dados, III) Síntese e IV) Análise. Primeiramenteé realizada a Extração de dados "brutos" de vários tipos de repositórios.…”

Section: Mineração De Repositórios De Softwareunclassified

CoDiVision: Uma Ferramenta para Apoio na Avaliação de Estudantes no Ensino de Programação.

Moura¹,

Ibiapina²,

Lira³

et al. 2017

Anais Do XXVIII Simpósio Brasileiro De Informática Na Educação (SBIE 2017)

View full text Add to dashboard Cite

Abstract. In this work is presented the CoDiVision tool, which was developed to infer the contribution level of developers in a software project. This tool can be used to support teachers in students evaluating in disciplines related to programming teaching. Through the use of this tool, instructors can evaluate the students individual contribution in works done in group, which are very common in the programming disciplines, since they allow to exercise several activities related to teamwork, but make the process of individual evaluation more complex. The CoDiVision was used in the evaluation of students in works developed in group, during a discipline designed to teach software development. Results show that the tool generates fundamental information to aid the students evaluation, especially with regard to notes attribution.Resumo. Neste trabalhoé apresentada a ferramenta CoDiVision, desenvolvida para inferir o nível de contribuição de desenvolvedores em um projeto de software. Tal ferramenta pode ser usada para apoiar professores a realizarem a avaliação de estudantes em disciplinas ligadas ao ensino de programação. Por meio da utilização dessa ferramenta, os instrutores podem avaliar a contribuição individual de estudantes em trabalhos feitos em grupo, que são bastante comuns em disciplinas de programação, pois permitem exercitar diversas atividades ligadas ao trabalho em equipe, porém, tornam o processo de avaliação individual mais complexo. A CoDiVision foi utilizada na avaliação de estudantes em trabalhos desenvolvidos em grupo, durante uma disciplina visando ensinar o desenvolvimento de software. Os resultados obtidos mostram que a ferramenta gera informações fundamentais para auxiliar na avaliação de estudantes, especialmente no que se refere a atribuição de notas.

show abstract

“…Some files may be considered defective while they are not and vice versa. Data collection and data cleansing are among the research topics that many mining software repository researchers are interested in Hemmati et al (2013b). In this paper, we use the above approach as a basic commonly used technique (D'Ambros et al, 2012;Hassan, 2009) for detecting fixes from version control systems, acknowledging that the accuracy of data collection can be improved by further research.…”

Section: Data Collectionmentioning

confidence: 99%