Collab-RS: semantic recommendation of external collaborators for projects in software ecosystems

The software development industry has evolved in recent years, presenting new challenges. In this scenario, software ecosystems have emerged as a new development paradigm through which external contributors support software production by providing solutions that complement a common ecosystem platform. Due to the many technologies, frameworks, and domains that an ecosystem can host, many collaborators acquainted with various domain topics and skills have also come into play. Recruiting collaborators becomes complex due to the varying degrees of knowledge and skills each collaborator has and their multiple competencies. There is a need to support the decision-making in the collaborator's recruitment, using the knowledge related to their skills. This work presents a solution supported by an ontology capable of recommending external collaborators for specific projects. The solution encompasses an architecture based on semantic models and expertise retrieval techniques. The architecture scores the collaborators' level of knowledge about topics and provides contextual information for the recommendation. Two studies were conducted involving two real software ecosystem platforms (Node.js and E-SECO). Results reveal that our approach can (i) use semantic models and inference mechanisms, (ii) offer context information essential for recruiter decision-making, and (iii) support recruiter' decision on contributor selection.