Abstract

This study aimed to investigate promising research areas in the field of social science by exploring the possibility of employing altmetrics to serve the purpose. To facilitate the research, five analytical methods were conducted: 1) topic modeling to derive research topics, 2) PCA to extract common features of bibliometric indicators and altmetrics indicators, 3) cross-correlation analysis and 4) regression analysis to identify time differences between bibliometric and altmetrics indicators, and define their relation, and 5) correlation analysis to finalize emerging research topics. The findings of the study revealed that the captures, among the altmetrics indicators, can be used as a leading indicator of the number of articles, authors, usages, and it can contribute to identifying emerging research topics two to three years earlier. This study has significance as it identified the applicability of altmetrics in overcoming the limitation of time lag that occurs in citation and keyword-based indicators used in previous studies, and proved it to be an indicator to ensure immediacy in the detection of emerging research topics. It also demonstrates the methodological extensibility for predicting emerging research topics by using PCA and provides the foundation for designing detection models for emerging research topics in the future.