Topic Mapping Approach To Automatic Classification Of Scientific Literature: A Study On Ecg Monitoring Systems

Abstract

With the significant growth in the number of research papers published in several fields, manual analysis of literature is becoming a resource-consuming and expensive process. This in turn accentuates the need for automated solutions. Hence, we propose an automatic classification approach to scientific literature based on factor analysis and topic enrichment techniques. This framework supports researchers in identifying the main themes and research topics in their areas of research early on before conducting an exhaustive, time-consuming manual review of the literature. The results of the proposed framework lead to a well-informed and focused manual review process later in the research cycle. We illustrated the proposed example with a working example focused on ECG monitoring systems. We further validate the proposed framework with a third-party literature classification tool as well as with an expert’s taxonomy. Validation reveals that the proposed framework generates thematic clusters highly correlated with research themes generated by the third-party tool as well as with the expert’s taxonomy © Technomathematics Research Foundation

Keywords: Automatic classification, Cluster, ECG monitoring systems, Factor analysis