Keywords: natural language processing, scientometric horizon scanning
Thorough assessments of evolving areas of interest in science and technology require scalable methods of scientometric horizon scanning to trace flows of ideas through vast collections of documents. This article investigates the scalability of a direct approach to scientometric horizon scanning which enlists methods of natural language processing (NLP) at low computational cost to trace usage, modification, and contextual association of noun phrases. Findings demonstrate the simplicity, efficacy, and scalability of this direct, NLP-assisted approach to scientometric horizon scanning.