Abstract
A novel bibliometric methodology based on natural language data processing for identifying emerging topics in science is presented. Along with the usual practice of data collection and preprocessing, our method includes a natural language processing (NLP) technique and an innovative mathematical function data approximation, which better addresses the data fluctuations problem along with allowing introduction of a specific measure of emergence, as well as to make predictions. It provides a systematic and data-driven approach to understanding the evolution of research topics and a better way to identify emerging topics in science.