AUTOMATIC TEXT CLASSIFICATION TO SUPPORT SYSTEMATIC REVIEWS IN MEDICINE

Medical publications continue to be a prolific source of information at an astronomical rate, with the amount of information added every day. Increasing amounts of information are being processed through systematic reviews. EBM relies on them as a fundamental tool. In this paper, we demonstrate that automatic text classification can speed up the reviewing process for medical topics. we propose a per-question classification strategy that exploits the specific protocol of a systematic review. This study shows that when integrating the classifier into the review workflow, the per-question method outperforms the global method. By locating and investigating the most recent applicable research, medical systematic reviews present thorough responses to specific concerns within an admittedly restricted field of expertise. A group of experts in the field must evaluate numerous articles as part of this procedure in order to discover the suitable instances, which typically takes a lot of time and effort. Encouraging this method is our aim. employing automated tools. A few weeks ago, it was trendy to use text methodologies for classification to partially automate the screening process by giving the panel of experts decisions support, thereby lowering the time and labour demands. In the present article, we make an impact on this field of research by performing a thorough set of categorization of text tests on a corpus created with an analysis, we make a contribution to this field of research in this paper.