Um modelo dinâmico de reputação para apoiar a manutenção colaborativa de software

Abstract

The importance of software in organizations is growing. However, to maintain its value, the software must be changed and updated. Software maintenance depends on the allocation of human resources to fulfill defined change activities. However, in a distributed scenario in which collaboration is critical to the well running of activities, designate developers for maintenance activities becomes a non-trivial task. In this context, reputation becomes a key element, affecting elements of collaboration, such as: coordination, cooperation, and communication. Therefore, tracking reputation evolution is important to promote collaboration in maintenance activities. The theory of System Dynamics can be applied in monitoring the evolution of reputation. Through the data obtained, it is possible to understand the past, establish what occurs in the present, and project future reputation behavior. Therefore, this work presents a model for calculating the reputation of software developers, supported by System Dynamics techniques, which allows simulating how reputation behaves over time. This model served as the basis for building an infrastructure for dynamic reputation information, which aims to enable the management and tracking reputation information of geographically distributed developers to support the allocation of these developers to maintenance tasks. In addition, it provides visualization and collaboration elements in an integrated environment for software maintenance activities. A proof of concept and an experiment made with real data of a company are presented with the intention of identifying the feasibility and adherence of the proposed model, as well as of the other resources offered by the infrastructure.