Towards a Maturity Model for Systematic Literature Review Process
Vinicius dos Santos, Rick Kazman, Rafael Capilla, Elisa Yumi Nakagawa
TL;DR
The paper addresses the lack of structured guidance to mature systematic literature review (SLR) practices in software engineering. It introduces MM4SLR, a CMMI-inspired maturity model built from 39 key practices organized into nine goals and five process areas, with five maturity levels, and validated via a four-SLR proof-of-concept. The model helps researchers plan, execute, document, and continuously improve SLR quality, enabling better reliability, replication, and updates. The authors discuss limitations and outline future work, including broader validation, clearer KP-to-level rules, and machine learning-assisted support for SLR tasks, aiming to raise overall research quality in software engineering literature reviews.
Abstract
Systematic literature reviews (SLR) have been increasingly conducted in software engineering and they provide significant benefits in terms of summarizing the state of the research. The process of conducting SLR is complex, involving several activities and consuming considerable effort and time from researchers. Researchers often skip or poorly conduct essential activities, which introduce threats to validity, resulting in lower-quality SLR. But researchers are often unaware of what they could do to mature their SLR process, thus improving the SLR quality. The main goal of this paper is to introduce a maturity model for the SLR process named MM4SLR. To this end, we were inspired by well-known models like CMMI (Capability Maturity Model Integration). We first identified 39 key practices for SLR from the literature and grouped them into nine goals that were further grouped into five process areas. We then organized the process areas into five maturity levels which compose our model. Our proof of concept, applying the MM4SLR to four published SLR showed that the MM4SLR is suitable for appraising SLR and can identify important flaws in SLR quality. MM4SLR can therefore support researchers in creating their SLR processes and selecting practices that could be adopted to mature their processes.