Uma arquitetura para avaliação de saúde de ecossistemas de software

Abstract

The complexity of the Software Ecosystems (SECO) environment, where process and project management are not trivial, brings major challenges to quality control. In addition, investments made by companies and external developers may be lost if ECOS expires. All these facts refer to the health of a SECO, which is the capacity to be maintained and expanded over time in the face of challenges. Based on this problem, the Heal Me architecture was proposed, whose objective is to automatically analyze the health of a SECO. In the context of this research, a systematic mapping was performed, through which the current solutions were identified, as well as a set of metrics for health evaluation. These metrics were evaluated through an observational study, formalized and automated as semantic rules. With these resources, a domain ontology was developed to analyze the SECO environment and the application of semantic rules for health analysis. Heal Me architecture was proposed and developed over the ontological base, a software architecture capable of capturing SECO data and performing health analyzes in an automated way. An important aspect of the automation is the use of APIs repositories for automatic data capturing. A case study was carried out with specialists to analyses the effectiveness of this first implementation effort. The results obtained point to the feasibility of the proposed architecture, as well as vulnerabilities to be treated in the next versions.