Uma Abordagem para Alocação de Equipes em Projetos de Desenvolvimento Distribuído de Software com base em Atributos Não-Técnicos

• Autor(a) principal: Ribeiro, Bruno Luna
• Data de Publicação: 2013
• Tipo de documento: Dissertação
• Idioma: por
• Título da fonte: Biblioteca Digital de Teses e Dissertações da UFPB
• Texto Completo: https://repositorio.ufpb.br/jspui/handle/tede/6097
• topic: Desenvolvimento distribuído de software; Alocação de equipes; Algoritimo genético; Distributed development of software; Allocation of teams; Genetic algorithm; CIENCIAS EXATAS E DA TERRA::CIENCIA DA COMPUTACAO

Taking into account the increasing adoption of projects and processes of Global Software Development (GSP), many organizations that practiced traditional software development, colocated, began to operate in the global market, dividing their collaborative units in many points of the globe. Besides the proximity to the international market, the use of GSD allows a reduction in time-to-market, improved quality of generated products, cost reduction and greater ease to find skilled labor. However, the distance between the teams involved in the GSD creates a variety of communication problems not faced in co-located development. According to NBR ISO 10006, human factor is the key to the success of software projects, since it is a genuinely human activity. In traditional development, companies use resources allocation processes to find the appropriate professional to perform a particular activity. However, geographical distance, and also the cultural and temporal factors existing in GSD are not classified in these processes. In this context, in order to reduce communication problems during the implementation phase, this work presents an approach for allocating teams in GSD projects, aiming to mitigate the problematic effects generated by the need for communication between geographically dispersed teams. Considering that the coupling between software components interferes to the communication required to develop them, the proposed approach takes into account non-technical characteristics of the teams such as temporal, cultural and affinity attributes, to discover the degree of proximity of the teams, comparing them with the level of coupling between components in order to identify which teams have better able to implement the software components with suitable quality of communication with other teams. It should be noted that the teams that are candidates for allocation based on non-technical attributes are technically qualified teams for the implementation of components. Due to the high complexity of the allocation problem, it is used a genetic algorithm with fitness evaluation based on the non-technical attributes for the allocation of the teams, making the problem computationally treatable in time.