A keyword-based approach to analyzing scientific research trends: ReRAM present and future

Research trend analysis is a primary step in defining research structures and predicting research directions from scientific papers. Recently, due to millions of annual scientific publications, researchers demand analytical methods to interpret the research field topologically and temporally. In this study, we propose a keyword-based research trend analysis method that automatically and systematically analyzes the research field by extracting keywords and constructing a keyword network. We verified our method on the resistive random-access memory (ReRAM) research field, which is in the limelight as an alternative device for non-volatile memory and artificial synapses. Our method performs three sequential processes: article collection, keyword extraction, and research structuring. We identified three keyword communities of ReRAM based on the processing-structure-property-performance (PSPP) relationship and found an upward trend in Neuromorphic applications. As a result, our method successfully structures the ReRAM research field and is expected to provide detailed insights into various research fields.