The main goal of a Systematic Review is to identify, evaluate, and summarize the findings of all relevant studies over a topic or an issue, making the evidence accessible to decision makers. But the process of manually conducting a systematic reviews takes a lot of time and researchers often have to limit their procedures. With the recent technological advantages, machine learning (ML) and text mining (TM) became useful to aid the systematic review process. The objective of this study is to detect the main trends of these disciplines by carrying out an analysis of a set of relevant articles, identified with a scientific database search between 2015 and 2020. Our analysis showed that mostly ML and TM techniques were applied to three steps: search, screening and data extraction. Huge progresses have been made over the years, but full automation remains a distant goal at present.