Abstract

Recommender Systems (RS) are increasingly applied to address diverse Software Engineering (SE) challenges, showing promise for enhancing various SE activities. The significance of this research lies in providing the status quo of the field to understand where research has been done, where there are deficiencies, and to observe trends. Current studies on RS solutions for SE tasks present a fragmented knowledge landscape. This disorganization makes it difficult to navigate existing research and understand the overall picture. Consequently, practitioners struggle to find suitable RS tools, and researchers find it hard to grasp the current state-of- the-art, pinpoint research gaps, or effectively position new contributions. This thesis aims to bridge this knowledge gap. It systematically identifies, synthesizes, and classifies the existing body of research concerning Recommender Systems within the Software Engineering domain. To achieve this, we conducted a Systematic Literature Review following established methodological guidelines. A taxonomy was then formed from the synthesized data to organize our findings. Our review successfully mapped the research landscape, highlighting dominant themes and common techniques. We also identified key gaps, particularly in software engineering task coverage and the adoption of certain RS methodologies. This work provides a solid foundation for practitioners evaluating RS solutions and offers clear directions for researchers. It aims to advance the frontier of the application of recommendation tools to enhance software engineering practices.