BDBComp
Parceria:
SBC
Mineração de regras para solução de problemas relacionados à fragmentação do Android

Marcus Adriano Ferreira PereiraMarcelo de A. Maia

Android operating system is highly fragmented resulted from a combinationof different versions of the operating system and thousands of differentdevices. Ideally, the developer is required to grasp the specific features of eachavailable device (e.g., sensors, graphical resources, among others) to ensurean adequate user experience. However, large fragmentation level makes it impossibleto systematically analyze all the possible options, which is responsiblefor failures in situations that occur only in specific versions of the API and/ordevice. An alternative for dealing with this problem is to rely on the crowd knowledgeto find possible corrections for typical failures linked to fragmentationproblems. This study aims to mine Android projects available on GitHub forassociations between elements of the Android API that co-occur in code to mitigatea fragmentation-related problem. We observed that some rules are mainlylinked to the user interface, and there is a concern of developers to use featuresof the new APIs, but also embrace users who use older device models but donot have enough resources for such. The rules could be used for recommendingpatches where the respective method calls occur.

http://www.lbd.dcc.ufmg.br/colecoes/vem/2017/013.pdf

Caso o link acima esteja inválido, faça uma busca pelo texto completo na Web: Buscar na Web

Biblioteca Digital Brasileira de Computação - Contato: bdbcomp@lbd.dcc.ufmg.br
     Mantida por:
LBD