Estudo, definição e implementação de um sistema de recomendação para priorizar os avisos gerados por ferramentas de análise estática
Carregando...
Data
2014-11-19
Título da Revista
ISSN da Revista
Título de Volume
Editor
Universidade Federal de Goiás
Resumo
Recommendation systems try to guide the user carrying out a task providing him with
useful information about it. Considering the context of software development, programs
are ever increasing, making it difficult to carry out a detailed verification of warnings
generated by automatic static analyzers. In this work, we propose a recommendation
system, called WarningsFIX, which aims at helping developers on handling the high
number of warnings reported by automatic static analyzers. The back end of this system is
composed of seven open-source static analysis tools collecting data, which subsequently
are used for visualizing information through TreeMaps. The intention is to combine the
outcomes of different static analyzers such that WarningsFIX recommends the analysis
of warnings with highest chance to be a true positive. Therefore, the information related
to warnings are displayed in four levels of detail: program, package, class, and line. The
nodes may be classified in the first three levels: amount of warnings, number of tools and
suspicions rate. An exploratory study was carried out and the limitations, advantages and
disadvantages of the proposed approach were discussed.
Descrição
Palavras-chave
Citação
MENDONÇA, V. R. L. de. Estudo, definição e implementação de um sistema de recomendação para priorizar os avisos gerados por ferramentas de análise estática. 2014. 109 f. Dissertação (Mestrado em Ciência da Computação) - Universidade Federal de Goiás, Goiânia, 2014.