| Popis: |
Background: Systematic literature review (SLR) has been widely adopted to synthesize evidence in a reliable and unbiased manner. The process of conducting SLR is rigorous and well-known; however, most SLR have not followed this process systematically, leading to various problems, including poor documentation, lack of quality, and difficulty to be reproduced. These problems are added to the inherent difficulties in conducting SLR, especially regarding effort and time consumption. Several practices (including methods, techniques, and tools) have already been experimented with to mitigate such problems. Still, it is unclear which practices should be prioritized to reduce such problems and difficulties. Aims: The main goal of this paper is to contribute to improving the SLR process being adopted by the community. For that, we introduce the idea of the Maturity Model for SLR process (MM4SLR). Method: We systematically examined the literature searching for practices destined to support or improve the SLR process. After collecting 84 practices, we analyzed, grouped, and synthesized them into 46 key practices further distributed into 15 specific goals. Inspired by the maturity model for software development process, we also propose five process areas and an initial version of MM4SLR containing four maturity levels. Results: MM4SLR provides a new view on those practices that could be progressively adopted to improve the SLR process. MM4SLR could define a pragmatic pathway that guides the adoption of those practices, which better fit the reality of who conducts SLR. Conclusion: The SLR process is very complex and should be more investigated. We argue that ways to better select and adopt the various existing practices are necessary. We conclude that a long research agenda to be fulfilled addressing the SLR process still exists. |
