Terminology Saturation Analysis: Refinements and Applications

In this paper, we outline the results of our recent research on terminology saturation analysis (TSA) in subject domain-bounded textual corpora. We present the developed TSA method. We further report about the two use cases that proved the validity, efficiency, and effectiveness of TSA. Based on our experience of TSA use, we analyse the shortcomings of the method and figure out the ways to refinement and improvement. Further, we share our prognoses on how TSA could be used for: (i) generating quality datasets of minimal size for training large language models for performing better in scientific domains; (ii) iteratively constructing domain ontologies and knowledge graphs that representatively describe a subject domain, or topic; or (iii) detecting and predicting events based on the TSA of textual streams data.