A ranking-based automated approach for supporting Literature Review research methodologies.
Literature reviews in general are methodologies of research which aim to gather and evaluate available evidence regarding a specific research topic. A common scientific method for performing this literature reviews is Systematic Literature Review (SLR). Another method is called Systematic mapping study (SMS). Their process if conducted manually can be very time and effort consuming. Therefore, multiple tools and approaches were proposed in order to facilitate several stages of this process. In this PhD thesis, we aim to evaluate the quality of these literature reviews studies using combined aspects. We measure the quality of the study`s included primary selected papers by combining social and academic Influence in a recursive way. Additionally, we will apply a machine learning ranking model based on a similarity function that is built upon bibliometrics and Altmetrics quality criteria and full text relevancy. In order to achieve the proposed approach, we begin with investigating the current state of the art in different directions, mainly the most effective and commonly used quality measures of publications, Altmetrics, Bibliometrics and machine learning text related techniques. A method for assessing the quality of these literature reviews research methods, would definitely be useful for the scientific research community in general, as It would save valuable time and reduce tremendous required effort.